Psychology, The Science of Behavior
1
What is Science
Science or scientific methods
consists of the methods for collecting, analyzing, and drawing conclusions from
data.
Assumptions of all Science
Basics assumptions of all
sciences are the same. We will discuss these assumptions one by one.
Behavior is determined
Believe that there are specific
causes for each of our behavior is called determination. Near about all
psychologists believe on determinism, may be form is different. Like according
to Freud, childhood sexual traumas are the cause of psychopathology, in
contrast according to the behaviorism, faulty learning is the cause of
psychopathology. Some people reject the idea of determinism. They believe that
soul and mind control person’s every behavior, so there is no cause of any
behavior. This thought is referred to as will.
We can measure the critical variables
It can do the study of the
behaviors or event which can not seen or handling. I.e through using the study
of behavioral events we can draw inferences about mental events. Like scientist
use the different observations in order to draw inferences about the stars,
planets and other unapproachable objects.
What is Scientific Research?
Here we will discuss a number of
characteristics which make the research scientific.
Empirical Analysis
It means that researchers use
observations and experimentations for learning.
Public Verification
It means that if any one use the
same procedure (psychological instrument, research design, sample etc ), he or
she will get same results like us. This phenomenon is called public
verifications. Every one can verify the results after using the same procedure.
This is also called the repetition of the research or experiment.
Systematic Observation
It means that we observe, study
or compare the person’s behavior through using different condition in
systematic way. The goal of systematic
observations to examine the phenomenon under as many relevant situations as
possible. We continue to repeat the experiments to determine which conditions
produce the effect and determine whether other factors affect the phenomenon.
Control of Environment
In all form of research,
researchers attempt to control of environment in some way. We use this
phenomena to protect our research to be a confound. Researchers have the greatest level of
control when they conduct research in a laboratory setting because they can
control many or all of the environmental conditions. There are many conditions also,
when researchers are not able to control the environment, i.e. in naturalistic
observation or in field study etc.
Rational Explanation
It means a sound explanation that
behavior is determined (cause & effect) and follow a lawful pattern. Some
people give irrational explanations about cause and effect, this is called
pseudo explanation. Two common names of pseudo explanations are nominal fallacy
and the circular or tautological explanation.
The Foundation of Research
2
The Hypothesis in Research
Hypothesis
It is a prediction about the
relation among two or more variables based on a theory or on previous research.
Where Hypothesis come from
Hypothesis come from theory or
previous researches. We read previous researches and check out the extent and
direction of relationship between variables, then we draw our own hypothesis
according to them.
Theory
Theory a set of general
statements and claims that allow us to explain and predict various events. For
example. This is a theory that aggression and violence are learned behaviors.
In theory there will be an explanation of this statement that why people behave
violently.
Characteristics of Hypothesis
Hypothesis direct our Observation
When a researcher prepare a
hypothesis. He or she will focus their attentions on the type of he information
that they will study. Hypothesis identify the variables those the researcher
will examine.
Important
Terminologies related to Research
Variable
A variable is any characteristic
of a person, non human subject, environment, or research condition that can
have different values.
Constant
A numerical value that ,by
definition, can not change. Like 1 hour always contains 60 minutes.
Independent Variable
IV is the type of variable which
use to predict, describe, explain or getting change in dependent variable.
Dependent Variable
It is condition that the
researcher wants to explain, predict or describe.
Manipulated Independent variable
Variable that researchers
manipulate or control referred to as manipulated independent variables.
Subject variable
The subject variable is a
condition that the researcher can not manipulate or control. Like personality
types, age, sex etc.
Hypothesis Describe the Relation among Variable
One of the most important
component of the variable that it describe the relationship between variables.
Like depression is negatively related with self esteem.
Hypothesis refer to the Population
Hypothesis also gives us the
knowledge of population of our interest. Like if I prepared a hypothesis that “there
would be a positive correlation between self esteem and academic achievement
among adolescents” so here every one can under stand that I am going to conduct
a research on “adolescents” population.
Types of Hypothesis
There are four types of
hypotheses. Keep in mind that types of the hypothesis based on the type of the
prediction which the researcher want to make.
Estimation of Population Characteristics
The goal of this type of
hypothesis is to estimate the characteristics of a population using the information
from a sample. Researchers use few terms for estimating the population
characteristics like sample, descriptive statistics, parameters etc.
Correlation among two
Variables
Second type of hypothesis state
the correlation between two variables. Correlation means some thing link
together. It means if one thing increase, other either increase or decrease. When
we hypothesis about correlation, we can predict that there will be a positive
or negative or no correlation between the two variables. ) In positive
correlation two variables tend to move in the same direction. When the X
variable increases, the Y variable also increases and if the X variable
decreases, the Y variable also decreases. ) In a negative correlation. The two
variables tend to go in opposite directions. As the X variable increase, the Y
variable decreases. That is, it is an inverse relationship. Numerically the
signs of + or – indicate the direction of the relationship between variables.
The correlation will be in between the range of +1.00 to -1.00.
Difference among two
or more population
Some time we want to estimate the difference among two or
more population. He we use independent variable, dependent variable or subject
variable. And t-test and F-test etc will be use for statistical analysis.
Cause and Effect
Its very popular, famous and
demanding hypothesis type. In this hypothesis researchers want to predict or
explain that change in dependent variable is only due to the change in
independent variable. They want to discover the cause and effect relationship
between variables.
Measurement
Measurement is a very famous
concept in every type of science and research. Measurement allows the
researchers to record and quantify our observation in an objective and
consistent manner. Measurement is objective, when does correctly and researcher
can verify previous findings of specific phenomena.
Features of all forms of measurements
Operational Definition
Operational definition is a rule;
we use to convert our observation into numbers.
Scales of measurement
Nominal Scale: A nominal scale consists of a set of categories that
have different names. Measurement on nominal scale label and categorize
observation, but do not make any quantitative distinctions between
observations. Although the categories on a nominal scale are not quantitative
values, they are occasionally represented by numbers. For examples room and
offices in the building may be represented with numbers. Like room no 1, room
no 2, room no 3 and so on.
Ordinal Scale: An ordinal scale consists of a set of categories
that are organize in an ordered sequence. Measurement on an ordinal scale rank
observations in terms of size or magnitude. Like! First position, second
position and third position, intense fear, low fear and minimal fear etc.
Interval & Ratio Scale: An interval scale consists of ordered
categories where all of the categories are intervals of exactly the same size.
With an interval scale, equal differences between numbers on the scale reflect
equal differences in magnitude and ratio scale having the additional feature of
absolute zero. For example! Zero is meaningful, 0 height means no height or
zero weight means no weight.
Reliability of Measurement
Reliability refers to the
consistency of measurement. A reliable test will produce similar result each
time we measure the same thing. Unfortunately no test available which is
perfectly reliable because measurement error effect all measurement techniques.
The classical theory or test reliability states that any test score is the sum
of the true score plus the effects of measurement error.
Test score = True
score ± Measurement Error
The symbol in the equation
indicates that measurement error can add to and remove from the true score of
the construct that we want to measure. One of the common methods to improve the
reliability of the data is to increase the number of measurement. Like good
carpenters always measure the things twice or thrice before cutting or
drilling.
Validity of Measurement
Validity refers to degree to
which a test or research method accurately measures the construct of interest. If
we say that test is not valid, we imply that the test measures things unrelated
to the construct we want to measure.
Types of Validity
Face Validity
Appearance of test is matter. If
a test is not authentic by face, it will be not valid. If the people taking the
test do not believe that the procedure is valid, they may not respond honestly.
Content Validity
Means that content of the
measurement tool should be the relevant to the targeted behavior (which
researcher wants to measure). For example if I am going to measure the
depression level of some one, definitely I will use the questionnaire; comprise
of the depression related questions.
Predictive Validity
Any good and valid test allow us
to predictive the other behaviors. For example test used in entry test predict
the student’s behavior that, what type of performance they show after admission
in any institution.
Concurrent Validity
Concurrent validity means test scores
should be relevant to the other similar behaviors. i.e if researcher administered
the self esteem scale on any person, then scores should be similar to the
scores find from self concept scale of similar person.
Survey of Empirical Methods
3
Internal and External Validity
Internal validity refers to the
cause and effect relation between the independent variable and the dependent
variable. An experiment is internally valid means that there is no doubt that
changes in the independent variable, which the experimenter controlled, caused
changes in the dependent variable.
Threats to Internal Validity
Any threat to internal validity
is a confounding variable. Confounding variable is an uncontrolled condition
related to the independent variable. Researcher can not control or manipulate
the confounding variable. There are four main categories of threats to internal
validity. Let me explain one by one.
Unintended sequence of events
The first threat to internal
validity is the unintended sequence of events. As the name implies, the data
may reflect a sequence of events that the researcher did not or could not
control.
Carryover effects
When one part of the study effect
the other part of the study. Or when researcher administered the same test for
two time. Participant’s experience after participation first time in the test
help him in second time. Which is major threat to the internal validity of a
research.
Maturation
If the research project covers
the long time. It is possible that the change in DV is due the long time and
person’s maturity. For example some depressed patients gradually improved with
our taking any treatment or therapy.
Intervening event
If the research covers long time,
participant may be experience any event which become the cause of the change in
dependent variable. For example during study participant lost his loved one,
which will draw an adverse effect on his mood, lead to adverse effects on
results of the study.
Non equivalent Groups
The second threat to internal
validity is the problem created by non equivalent groups. When we want to study
the effects of independent variable on dependent variable. The both groups
should be equal except independent variable. In the case of non equivalent groups, research could not be
able to draw the conclusion that either change in DV is due to the IV or the
other confounding variable. Like age, gender or sex etc.
Mortality
Mortality does not solely means
the death of the participants, it refers to the termination of the participants
from research study during continuation of the study. It may be due to the participant’s death, bored ness or migration
from one city to an other city or country.
Subject Variable
For many research projects, the
independent variable is a subject variable. Because the researcher does not
have direct control over this variable, he or she can examine the relation
between the independent and depend variable, but can not automatically assume
cause and effect.
Measurement Error
This threat is related to the
measuring instrument, which is not valid for measurement of any construct.
Ceiling and Floor effects
Some time researcher may select
very easy test, which attempt every one, or some time very difficult test,
which is not in the approach of any one. In this way researcher will be not
able to find the differences among the participant’s score.
Reliability
Due to the invalid test,
reliability of the score will be also not sufficient.
Regression to the Mean.
Change in score represent the regression
to the mean phenomena. This effect occurs when people receive exceptionally
high or low scores on a test or measure. When tested a second time, the scores
will end to be close to the average. For example some people may receive a poor
score on a test because they do not feel well or for reasons unrelated to their
knowledge of the material. When we test them again a few weeks later, their
scores will probably be close to normal consequently, the change in test scores
may reflect regression to the mean and not the effects of the independent
variable.
Ambiguity of cause and effect
The final threat to internal
validity refers to the assumptions of cause and effect. We must show the cause
occur before the effect.
External Validity
External validity refers to the
generalization drawn from the data. it is related to the representative sample.
Sample will be more representative, research will be more externally valid.
There are two types of external validity. 1) Generality of finding refers to
generalization from sample to population. 2) Generality of conclusion refers to
the generalization from population to population.
Threats to External validity
There are three major threats to
external validity for any research. 1) recruitment of participants, 2) effects
of situation, and 3) effects of history. Let me elaborate these.
Recruitment of Participant
What is our method of selecting
participants for our research project. Either we are going to use random
sampling or convenient. Some time our method of recruitment of sample does not
allow us to make generalization of our results on large population. For example
we are going to find achievement level of adolescents, and we are going to draw
sample from university, is it possible that university adolescents sample will
represent the entire population of adolescence who work in shops, hotels or
related to the other type of occupations.
Effects of Situation
Situation is also matter in
research. We can not generalize the findings of one situation to another situation.
For example I can not generalize the findings of my university to the Oxford University
Effects of History
Can the results of research
conducting before 50 years generalize today? Time and situations matter in this
long history. Before 50 years female was not more efficient in mathematics than
today. Today they are provided more facilities to learn math.
Survey of Empirical Methods
In this section of the notes, we
will discuss some important methods of conducting research in science,
specially in social sciences. Let me explain gradually.
True Experiments
This method use for determining
the cause and effect. True experiment is very useful due to his quality of
control and remove the threats to internal validity, which are very important
for any successful and good research. It ensure that IV comes first and DV
comes later and it is also ensure that IV and DV are correlated. In this method
researcher can manipulate and control the IV and researcher use random assignment
for selecting the participants to the conditions.
Random Assignment
Random assignment means that each
participant selected for the experiment as an equal chance of experiencing the
one of the research condition.
Use of Control group
A control is the group of
participants who experience the same treatment as the other participants except
for exposure to the independent variable.
Intact Group Design
This method use when researcher
can not manipulate or control IV and use subject variable. In this method
researcher does not select the sample from a single population, several
population use for sampling which represent their subject variable. For example
patients of different disorders. Here we can not use random assignment due to
taking subject variable. Here we collect data and compare the difference among
groups.
Quasi Experimental Design
A quasi experiment is similar to
a true experiment except that we can not randomly assign the participant to the
participant to the treatment and control condition. The advantage of this
design to compare the groups. Like male vs. female etc. limitation of this
method that it can not explore the cause and effect. Here third variable
problem is also important. For example we don’t know that change occur due to
IV or any other variable.
Survey
Researchers use survey to
estimate the population parameter. It uses of find the opinions and attitude of
the population toward a specific concept, problem, issue or phenomena. The
advantages of the survey are ! it is not expensive, done quickly and provide a
large number of data from the population.
Correlational Studies
Correlation is used to measure
and describe a relationship between two variables. Usually the two variables
are simply observed as they exist naturally in the environment. There is no
attempt to control or manipulate the variables. Correlation can be classified
into two basic categories. Positive and negative correlation. 1) In positive
correlation two variables tend to move in the same direction. When the X
variable increases, the Y variable also increases and if the X variable
decreases, the Y variable also decreases. 2) In a negative correlation. The two
variables tend to go in opposite directions. As the X variable increase, the Y
variable decreases. That is, it is an inverse relationship. Numerically the
signs of + or – indicate the direction of the relationship between variables.
Range of the relationship between two variables will be between +1.00 to -1.00.
This is called coefficient of
correlation”. Co relational studies help the researchers to better understand
the complex human behaviors and again it is not able to find out the cause and
effect between variables.
Single Participant Research Methods
Single Participant experiments
This method use one participant.
Mostly use in experimental researches in order to determine the IV and DV
relationship and determine the cause and effect between variables.
Case Study
This is one of the most familiar methods
of single participant research method. In this method researchers study a
single participant in deep. The essential feature of this method is intense
review of the behavior of a person. The advantage of this method is that we can
use sound experimental techniques to determine how to influence a person’s
behavior. The disadvantage of this method is that researchers can not
generalize the results of a single person to the whole population.
A collection of statistical
techniques used to combine the results of separate research projects to help
one determine the relation between the independent and dependent variables. This
method used widely in psychology, sociology, political science, medicines and
other researches.
Learning About Psychology and Forming Hypotheses
4
Bibliographic Research
Three are three important levels
of bibliographic sources. We will discuss these sources one by one.
Tertiary Bibliographic Sources
This is third level source.
Textbook, articles, news papers, psychology dictionaries, encyclopedias are all
tertiary sources. Author of these sources did not review the topics in detail.
The advantage of a text book is that it gives you a quick and easy to
understand introduction to a topic. This advantage is also limitation.
Secondary Bibliographic Sources
These sources more emphasis on
the topic than tertiary sources. These are the comprehensive reviews and author
discuss on the topic in details. Text books having many chapters on a single
topic, discuss its current trend and theories are also include in secondary
sources. It also give the detail of method of conducting research on the
specific phenomena. This source emphasis on different issues related to the
research topic.
Literature Review
This is not secondary source but
this shares some features with secondary source. Literature review found in
professional scientific journals. Literature Reviews are more topic specific
than text books. It is also provide a comprehensive summary of current
research. Different journals are the best sources of getting literature review.
Primary Bibliographic Sources
Original research reports, you
can find in different journals are considered as primary bibliographic sources
of research. The advantage of the primary source is that there is no filtering
of information. These research reports often difficult for the readers, authors
of research reports of ten use these reports to communicate their research
conclusion to the other researchers.
Peer Review
This is the very good
characteristic of research journals and text books. The goal of the peer review
is to filter out bad research and make the good research as clear and
compelling as possible. Peer review is a part of the normal edition process
during which experts read a manuscript to ensure that the researchers used
appropriate methods to collect and analyze the data and then made reasonable
inferences from the data. when a researcher submit his or her research report
to the management of any journal, a group of the research mostly three to four researches,
having enough knowledge on the relevant topic discuss and review the research
report. After reviewing either may be the accepted, sending the report to the
author for amendment or rejected.
Developing a Search Strategy
Searching the Literature: The Library
A library is a wonderful place to
find information. How you can use your library. We will help you. Let me
explain !
Library Staff
It’s a very common thing, that
student ignore the very important aspect of searching from the library, it is
library staff. Librarians have the knowledge about the storage of books that a
book kept in which cupboard? Almost staff of every library know very well this
type of information.
Search Tool
Card catalogue and different
reader’s guides are the helpful for searching any book, journal or any other
source of searching literature.
Online Catalog
All libraries keep their
bibliographic material in a well organized system that you can use to find
information quickly. Specifically, most colleges use the library congress
classification scheme that uses a combination of letters and numbers to classify
each book, document, other materials such as films and records.
Sampling, The First Steps in Research
6
The nature of Sample
A number of ways that define
samples. First, sample is a valid representation of a population. Second way is
that compare the sample with those things, those are not sample. A true, valid
and successful research depends on the nature or type of sample. Type of sample
depends on that what type of method is used for data collection. Now we discuss
the type of the sample.
Scientific samples versus Anecdotal Evidence
Anecdotal evidence means
collection of personal experiences haphazardly that may be biased. For example
I went to Lahore ,
stay there just for 2 hours and came back. Here I draw inference that people of
Lahore are very
cruel. Is it true ? no its wrong, because this inference is depend on my
personal experience. May be I will met there just 10 to 15 people, those are
not the true representative of 500000 people of Lahore . And a scientific sample means that
the sample select by using true and good sampling technique and sample should
be the true representative of the population.
Probability Sampling
Probability sampling means that every one from the
population has an equal change of being selected for a research study.
Some important terms related to sampling
Sampling population
Selected group of people, who share the same characteristics
as the target population.
Subject Pool
The subject pool refers to a group of people to whom the
researcher has an easy access.
Sampling Frame
A list of names that identities the members of the
population is called sampling frame.
Characteristics of Probability Sampling
- Treat each member of the population as a potential member of a sample.
- We can determine the probability of selecting any specific sample. For example you have a population of 6 participants and you choose 2 of them. You will draw 15 potential samples (pair of two people). For example you have 1, 2, 3, 4, 5, 6. how much pair is possible (1, 2) (1, 3) (1, 4) (1, 5) (1, 6) (2, 3) (2, 4) (2, 5) (2, 6) (3, 4) (3, 5) (3, 6) (4, 5) (4, 6) (5, 6). And now you can randomly select any pair (sample of two participants).
Sampling Methods
There are many methods of drawing
samples from the population. Each method shares a common goal, ensure that the
sample is an unbiased depiction of the population.
Simple Random Sampling
Simple random sampling is occur
when each member of the population has an equal probability of selection. In
this method first researcher estimate the size of population then determine the
random number for selecting the participant in research.
Sequential Sampling
In sequential sampling,
researcher list the member of the sampling population and then select each kth
member from the list by using interval. Interval can be determine by using this
method
Sampling Interval =
Size of population ⁄ desired size of
sample
If researcher got 3 answer, then
it means that every 3rd person will be selected for sample.
Stratified Sampling
This technique is especially
useful in situations where the population contains different subgroups that the
sample must include. For example age, education, socioeconomic status etc. for
selecting the sample, in initial step researcher identify the subgroups and
then use simple random sampling or sequential sampling in subgroups for draw
sample.
Cluster sampling
This method use when it is
impossible to take one or two people out of their environment or when other
sampling methods are not possible or impractical. Researchers often use this
technique when working with intact groups. For example researcher is interested to find
the impact of teaching style on the students. There are 20 different sections
of the same subject of psychology. Here researcher use cluster sampling
technique.
Non Probability Sampling
In many situations researcher
will resort to non probability sampling. Using a non probability sampling
method may limit he inferences.
Convenience Sampling
Convenient sampling means that
the researcher uses members of the population who are easy to find. Interviewing
people in shopping malls are the example of convenient sampling.
Snow ball sampling
Some time it is not possible or
difficult for the researcher to find the members of a coherent group for the
research. In such situations, the researcher may wish to use snow ball
sampling. The researcher needs to find a member of the cohort and use him or
her to find other member of the cohort.
CREATING AND USING TESTS, SURVEYS, AND OBJECTIVE MEASURES
7
Measurement is the heart of
empirical sciences. There are two goals of measurement 1) for replacing the
ambiguity of words with operationally defined constructs. For example
operational definition of depression may be lack of sleep, lack of sexual
activity, sadness ete. 2) Second goal is standardization and consistency of
measurement. Consistency means compare people using a common set of procedures
and scales. Standardization means that number have constant meaning for every
one for every culture. Like 110 will be same in Pakistan , India , Germany and
Switzerland etc.
There are two common
misinterpretations of psychological tests and measurement.
Human behavior can’t be measure (
human behaviors are too complex to measure objectively. This is common complain
because some people think that Psychology is not a science.
Uncritical trust on numbers. This
problem is called mystique of quantity which means an exaggerated regard for
the significance of measurement, just because it is quantitative with out
regard to what has been measures (behaviors).
Creating a Measurement Scale
What questions are you trying to answer ?
Best way to answer this question
is to read the hypothesis. Well formulated hypothesis make you able to define IV and DVs. For
example if a researcher formulated the hypothesis that Depression is negatively
correlated with self esteem. It means that researcher want to calculate the
relationship between depression or self esteem (either increase or decrease in
depression draw effect on self esteem). Identifying the variables is half the
battle as it helps you identify the types of behaviors you want to assess.
What is the most convenient method for producing data
It depends on the nature of iv
and dv. For example if researcher want to check out the children’s behavior of
different ethnic groups, you will have to use observational technique for data
collection but if you want to compute the relationship between self esteem and
depression of the participants, you will use questionnaire for data collection.
What is most accurate measurement technique?
There is no perfect measurement
technique because all techniques have some sort of errors. Well you should use
methods which reduce the bias. For example for observational method you should
use some trained observers who observe and save all necessary and relevant
samples of behaviors which is needed.
Constructing Interviews, Questionnaires and Attitude Surveys
There are many ways that we can ask people questions like
face to face interviews to questionnaires. We will discuss one by one.
Personal Interviews:
Personal interviews are either
face to face or telephonic. It is very popular technique for gathering data
regarding person’s attitudes and beliefs. Specially face to face interviews
have high level of cooperation. Interviews can ensure that participants
understand the questions. Interviews can highly structures or unstructured may
be individual or may be in group setting. Along with advantages interviews have
also some disadvantages. For example its costly and time consuming, conducting
interview require much practice. Conducting interviews difficult when there are
cultural, language and racial barriers between interviewers and participants.
Self Administered
Surveys
Alternative of personal interview
is self administered surveys. Give many people questionnaires and receive back.
Advantages are: it is not costly, no required any special training, not time
consuming. (Use a captive audience) Use
available participants for getting information through surveys. For example
present students in introductory psychology class. It is very quick and easy
approach method. (Use social psychology
to your advantage) You can also use social psychology methods for
completion your surveys. It gives you the methods that how you can collect most
filled questionnaires form the participants.
Preparing the questions
When you select the method for
obtaining data, you have needed to prepare the questions. You can ask two types
of question 1) open ended 2) close ended.
Open ended questions (OEQ)
It gives the opportunity to
person to give short answer to the question.
For example
Do you satisfy with your
statistics teacher? In the answer of this question respondent can give the
short answer
Closed response questions (CRQ)
In CRQ researcher supply response
option to the participants. Researchers use many alternative formats for CRQ.
Nominal Category Response
I am: Single Married Divorced Widowed
Forced Choice Response
Participants asked to select the
response that best represents their answer to the question. For example
Shy vs Out
going
Leader vs Follower
Introverted vs Extroverted
Likert Format
It is one of the most popular CRQ
format. it offers a clear and unambiguous ordinal scale of measurement. You can
use same format for many questions.
Likert scale format consist of 4 to 10 points that represent potential reaction to the question or
statement. There will be a middle neutral option in the likert scale format.
for example
Overall, I believe that Chief Minister Shehbaz Sharif is doing a good
job
Strongly agree Agree Neutral Disagree
Strongly disagree
Guttman Format
A guttman format like the likert
format, represents a continuum that ranges from one extreme to another.
For this format researcher arrange the answer options by levels of acceptance. For
example
Which of the following best
represents your opinion regarding abortion? Select only one statement
- Abortion may be performed when the mother’s life is in danger.
- Abortion may be performed if the fetus has a severe disability
- Abortion may be performed in case of rape
- Abortion may be performed as a type of family planning
- Abortion may be performed for any reason.
Writing Good Questionnaire and Survey Items
Ask Single questions
or make single statement
Example: Do you agree with the college’s plan to build a new
student recreation center and increase the student fee to 500 dollars to help
pay for the building ?
Alternative: Do you agree with the college’s plan to build a
new students recreation center ?
Ask Specific
Questions and Avoid vague terms
Example: Do you actively support the college’s decision to
build a new student recreation center.
Alternative: I believe that the college should build a new
student recreation center as soon as possible
Write Neutral
Statement and Avoid a Biased Tone
Example: Many people believe that it is bad to spank children.
Do you ?
Alternative: Do you believe that parents should spank their
children.
Ask questions that
Don’t embarrass or Anger the Participant
Example: Are you still a virgin?
Alternative: With how many people have you had sexual intercourse?
Whenever possible,
Use Simple Words and Concepts
Example: Do you
believe that you received sufficient time to consider your alternative before
you were required to purchase your computer?
Alternative: Did you feel rushed to buy the computer?
Recognize one question
may not be enough
Example: All in all, I am inclined to feel that I am a
failure?
Alternative: I feel that I have a number of good qualities.
I am that I have a number of good
qualities.
I feel that I do not have much to be
proud of.
RESEARCH DESIGN I:
9
(Between Subject Design)
A research design refers to the methods used to collect data
that will decisively answer an empirical question. A good research design must
have three characteristics. 1) It must be efficient (design should produce the
best quality data in least amount of time and money). 2) The design must
eliminate factors those biased the results. 3) Design should produce the data
which clearly address the research question.
A MODEL FOR RESEARCH DESIGN
The purpose of any research
design is to protect researchers from lapses in objectivity and the bias inherent in
personal judgment. Especially in Psychology the participants come to us
with their anxieties, fears, biases and awareness that they are coming for
participating in research study. Good behavioral scientists will be those who
collect very useful data from those participants.
Purpose of Research
Here the first thing is the
personal bias of the researcher. Mostly research questions based on
researchers own experience. Every one explains the phenomena according to their
own self. The influence of researcher’s orientation can have considerable
influence on how he or she will view a problem.
Design of Research (control for extraneous variables)
After identifying the purpose and
the hypothesis, the next step is the design phase during which the researcher
develops controls for potential confounding variables. Researcher
formulates a design that what are the extraneous variables and how he or she
can control them during study.
Selection of Participants
This is the stage of selecting
the representative sample for the research. Most of the problems and the
threats to the internal validity occur due to the non representative samples,
which do not produce the true data for making inferences for the population.
Preparing the Participant for the Study
Independent Variable
Variable which use the researcher
for making prediction about independent variable?
Type of Independent variable
Type of independent variable is
draw a sound effect on the design of the research. It is also important in the
context of control conditions. There are
two type of IV (subject vs. manipulated). subject variable is the variable
which exist before the participation of the person in the study, like age, sex,
personality traits etc. manipulated variables are the variables which
researcher manipulate on the participants, like researcher select the alcohol
quantity for administering on the participants.
Between Subject versus within subject variables
We can also make a comparison of
IV in the context of between subjects or within subject variables. In between subject variable, participants
tested only under in one treatment or condition and in within subject variable,
participants tested under more than one treatment conditions. When we compare
the depression level among genders then it means that gender (male &
female) are the between subject variable because male type of sex did not give
the correct estimate of depression level of female gender and vise versa so the
sex is representative of between subject variable. On the other hand when researcher
observes the behavior of the same person under different treatment conditions
or level of IV so it is representative of with in subject variable. For example
when researcher observe the behavior of a person under different time table.
Dependent Variable
Dependent variable is the variable
which changes due to the change in independent variable. Researcher can
measure the dependent variable by evaluating the construct of dependent
variable. Like if researcher is interested in computing the effect of self
esteem on depression so in this case depression is an independent variable and
researcher will measure the depression by evaluating its contents for example
lack of sleep, lack of hunger, low interest in sexual activities etc.
Reliability and Validity
Reliability means the consistency
of scores and validity means the use of true instrument for measuring the
construct. Low reliability and validity of research projects is due to the use
of techniques that were unreliable, invalid or both. There are many ways to
ensure the reliability and validity of a measure. One is to use techniques that
other researchers use.
Objective Recording of Data
One important and useful method
for reducing experimenter bias single blind or double blind methods. The single
blind procedure indicates that the person participating in the research does
not know the specifies of the researcher’s hypothesis. The goal of using single
blind procedure is to reduce the effect of demand characteristics of the
participants. Demand characteristics means that participants having the
knowledge of the participants hypothesis and then participants behave according
to the hypothesis. In double blind studies participants and as well as
researchers did not know about the
research objectives.
Placebo
Placebo is a common technique
used to create a single blind. For example when researcher want to examine the
effect of a drug, he can use placebo along with real drug. Participants of the
research did not know that either they receive real drug or fake pill
(placebo). Placebo is also use in control group.
Carry over effect
Carry over effect refers to the
effect that previous experience have on subsequent behavior. As an example, if
researcher uses the same test on two occasions, the participants may remember
the questions and their answers, and this memory may bias the results.
Research Hypothesis
Hypothesis is the tentative
explanation of prediction of the relationship between two variables. Preparing
hypothesis is very important part of any research design. Hypothesis not only shape the
design of the research but effect on statistical analysis of the data.
Hypothesis tells the relationship between the groups. This relationship may be
directional or non directional. In this context research divide the hypothesis
into two types. 1) Directional hypothesis 2) Non directional hypothesis.
In directional hypothesis
researcher indicate the relationship between groups in the form of greater than
or less than. For example Women are more competent in mathematics as compare to
men. In non directional hypothesis researcher indicate the relationship between
groups but no direction of the relationship. For example there would be a
relationship between depression and self esteem among college students.
Mathematical Hypothesis
When researcher use the
mathematical impression in writhing the hypothesis, this form of
hypothesis is called mathematical hypothesis. When researcher uses the
mathematical impression for of hypothesis, they use two types of hypothesis 1)
first is Null hypothesis and second is Alternative Hypothesis. Alternative
hypothesis is also called research hypothesis or scientific hypothesis. Null
hypothesis indicates that there would be a no difference between populations
and alternative or research hypothesis means that there would be a difference
between groups.
Evaluating Hypothesis
When researcher use data for make
prediction, they use the term rejecting or fail to reject of null hypothesis.
When researcher make inference, they keep in mind the probability of doing
mistakes in rejecting or fail to reject the null hypothesis. There are two
types of mistakes or errors. 1) Type 1 error 2) Type 2 error. Type 1 error is
that when researcher reject the null hypothesis when it is true. This is due to
the mislead of data. It is due to the extreme samples and other random events.
For example difference between the groups. Type 2 error is when researcher is
fail to reject the null hypothesis when it is truly wrong or rejected.
Research Design II
Single Variable Between-Subjects Research Design
This research design is related
to single independent variable with more than two conditions. For this design
researchers use ANOVA .ANOVA allows us to examine the effect of many different
levels of the same independent variable on the dependent variable.
Research Design III
Between-Subject Factorial Designs
This research design incorporate
more than one independent variables. This design has the following advantages
1) examine the effect of more than one independent variable 2) examine the
interaction between independent variables. Two variable design examines how
much each independent variable, by itself, influences the dependent variable.
In addition ,the test determines how much the combination of the independent
variables affects the dependent variable.
2×2
It means that two
variables and two two conditions in each variables
Main Effects
A main effect refers to the effect that one independent
variable has on the dependent variable holding the effects to the other
variable constant.
Research Design IV
Correlated-Groups Designs
Repeated measure designs or
matched groups designs are called correlated groups designs. These designs
allow the researcher to exercise greater control over the variables that
influence the dependent variable. The main characteristic of this design
distinguished it from other designs is that it has some known or intentional
correlation among the variables. In this design researcher intentionally create
the groups in such a way that there is a correlation among pairs of scores
across the groups. Two primary techniques that researcher use is repeated
measure design or matched group design. Both are different to each other.
The logic of using these
techniques is to minimize the individual differences among participants which
play a very vital role in within group variance. There are several ways we can
reduce the within groups variance like
selection of more homogenous sample, second is to treat more significant
subject variable sex as one of the independent variable in a factorial design.
Another way to reduce the within groups variance is to use the correlated
groups design. This procedure is an exceptionally cost effective method for
increasing statistical power,. Like the factorial design, the correlated groups
design allows the researcher to reduce the within groups variance.
Repeated Measure Design
In the repeated measures design,
we collect the data from the same participants on a series of occasions or
under different levels of the independent variable to determine if there are
systematic changes in the participant’s behavior. There are two types of
repeated measure design.
The first type of repeated
measure design allows us to test the same participant using several levels of
the same IV. The advantage to this procedure is that it requires fewer
participants because we test each participant using each treatment rather than
assigning separate groups. Another advantage to this design is that we can use
each participant as his or her own control condition and thereby reduce the
size of the within groups variance in the ANOVA. The second type of repeated
measure design allows us toe use time as an independent variable. For this type
of research, we arrange to test or sample the participant’s behavior changes as
a function of time. This technique is useful when time is a critical predictor
variable.
Grounded Theory
Basic Principles of Grounded Theory
Building Blocks
Grounded theory as method
provides us with guidelines on how to identify categories, how to make links
between categories and how to establish relationships between them. Grounded
theory as theory is the end product of this process: it provides us with an explanatory framework
with which to understand the phenomenon under investigation.
Categories
These designate the grouping
together of instances for example events etc. that share central features of
characteristics with one another. For example anxiety, anger and pity under the
heading of emotions.
Coding
This is the process by which
categories are identified. In the early stages analysis, coding is largely
descriptive. Here descriptive labels are attached to discrete instances of
phenomena.
Constant comparative analysis
This ensures that the coding
process maintains its momentum by moving back and forth between the
identification of similarities among and differences between emerging
categories. Having identified and common feature that unites instances of p
phenomenon, the researcher needs to refocus on differences within a category in
order to be able to identify any emerging subcategories.
Negative Case analysis
This ensures that the researcher
continues to develop the emerging theory in the light of the evidence. Having
identified a category, or a linkage between categories, grounded theory
researcher need to look for negative cases that is, instances that do not fit.
The identification of such instances allows the researcher to qualify and
elaborate the emerging theory, adding depth and density to it, so that it is
able to capture the full complexity of the data on which it is base.
Theoretical sensitivity
This is what moves the researcher
from a descriptive to an analytic level. In grounded theory, the researcher
interacts with the data. That is , he or she asks questions of the data, which
are in turn modified by the emerging answers. Each emerging category, idea,
concept or linkage informs a new look at the data to elaborate or modify the
original construct
Theoretical sampling
This involves collecting further
data in the light of categories that have emerged from earlier stages of data
analysis. Theoretical sampling means checking emerging theory against reality
by sampling incidents that may challenge or elaborate its developing claims.
While the earlier stage of grounded theory require maximum openness and
flexibility to identify a wide range of predominantly.
Theoretical saturation
Ideally, the process of data
collection and data analysis in grounded theory continues until theoretical
saturation has been achieved. In other words, the researcher continues to
sample and code data until no new categories can be identified, and until new
instances of variation for existing categories have ceased to emerge.
Memo writing
This is an important part of
grounded theory. Throughout the process of data collection and analysis, the
researcher maintains a written record of theory development. This means writing
definitions of categories and justifying labels chosen for them, tracing their
emergent relationships with one another, and keeping a record of the
progressive integration of higher and lower level categories. Memos will also
show up changes of direction in the analytic process and emerging perspective,
as well as provide reflections on the adequacy of the research question.
Memory Work
Data Collection and analysis
This method involves a process of
group work during which co researcher remember, compare, discuss and theorize.
Memory work research can take many months or even years to complete, and it can
generate diverse insight.
Phase: 1 Generating memories
Step 1: Form a memory
work group
Memory work groups can have
between four and eight members. Such group share characteristics which may be
regarded as relevant in relation to the topic under investigation. For example
with topics such as emotions or sexual practices., single sex groups may be
preferable. For some topics, a group of friends may be best, whereas other
issues may be best discussed within an atmosphere of relative anonymity. The
most important requirement, is mutual trust and a feeling of safety within the
group.
Step 2 Choose a trigger
To prompt group members to write
memories, the group needs to agree on a suitable trigger. The trigger is a word
or a short phrase which is expected to generate memories that are relevant to
the topic under investigation.
Step 3 Write memories
Each member of the group writes a
memory in response to the trigger. The writing of memories can take place
either within the context of the group meeting itself or outside of it. To ensure
that the memories are written in a way which renders them suitable for memory
work analysis, they should be written in the third person singular and include
as much circumstantial detail as possible. This could be references to sounds,
tastes and smells that appear to be inconsequential ofr trivial in terms of the
event being remembered. It is important to avoid censoring apparently
irrelevant information and to avoid interpretation or justification. The style
of writing should be purely descriptive biographical.
Phase 2: Analysis of the memories
Step 4: Texture analysis
The memory work grouper convenes,
with members bringing along their written memories. Initially, each individual
memory is analyzed separately. Each member of the group is handed a typed copy
of each memory that has been produced. The memories are then examined, one by
one, in terms of the sequences of actions, role relations, clichés and contradictions,
statements made and absences which characterized them.
Step 5 Cross sectional analysis
Here the co researcher compare
their memories with one another. They look for similarities, differences,
recurring themes and common patterns.
Step 6 Re writing the memories
At this point in the process of
analysis, it may be helpful for group members to return to their memories and
to re write them in the light of the textual and cross sectional analysis. Re
writing memories can be a way of becoming more acutely aware of the pro esses
of meaning construction because it requires us to thing of different ways of
making sense of the same event. This is where memory work connects with
consciousness raising.
Phase 3: Integration and theory building
Step 7: Analysis of transcripts and memories
Now it is time to compare and
contrast the memories produced in response to several trigger. The transcripts
of the group discussions of the memories constitute further data for integrative
analysis. The ideas generated by the group in relation ot particular social
constructions are themselves subjected to critical appraisal and further
theorizing.
Step 8: Writing memory work
The process of writing up memory
work is still part of phase 3. writing both requires and facilitates
integration of material it is also likely that theoretical formulations will be
clarified, and even revised, during the process
of writing. Since memory work is a collective enterprise, writing can
not be separated from further discussion. Drafts of papers will have to be read
and discussed by the co researchers, and this may generate new ideas and
changes to the original analysis.
Interpretative Phenomenological Analysis
IPA works with transcripts of
semi structured interviews. Smith provides guidance on how to conduct
interviews hat will generate data suitable for phenomenological analysis. In
general, interviewing for IPA shares the principles and practices associated
with semi structured interviewing. Since phenomenological research requires the
researcher to enter the life world of the research participant, it is extremely
important that the questions posed to the participant are open ended and non
directive. Their sole purpose is to provide participants with an opportunity to
share their personal experience of the phenomenon under investigation with the
researcher. Focused and specific questions should be used toe encourage participant
to elaborate rather than to check whether they agree or disagree with particular
claims or statements. Diaries, video or written means also use for the data
collection for IPA. IPA use ideographic approach. One this is more important in
the IPA, that researcher’s point of view about the phenomenon is also include
in the interpretation.
The first stage
of analysis in IPA involves reading and re-reading of the text. Researcher produce wide ranging
notes that reflect the initial thoughts and observations he or she may wish to
record in response to the text. These could include association, questions,
summary statement and open coding like grounded theory.
- Identification
of themes (each section)
The second stage
of analysis requires the researcher to identify and label themes that
characterize each section of the text. Themes titles are conceptual and they
should capture something about the essential quality of what is represented by
the text. Psychological terminology may be used at this stage.
- Clustering the
themes
The researcher
listed the themes identified in stage two and thinks about them in relations to
one another. Some of the themes will form natural clusters of concepts that
share meanings or references, whereas others will be characterized by
hierarchical relationships with one another. For example themes such as
childhood memories, going to school and relationship with mother could form a
childhood cluster.
- Production of
a summary table
The forth stage
of analysis production the summary table of structured themes, together with quotations
that illustrate each theme. The summary table should only include those themes
that capture something about the quality of the participant’s experience of the
phenomenon under investigation. This means that some of the themes formulated
in the stage 2 will be not included in the summary.
Discourse Analysis
- Discursive constructions
The first stage
of analysis is concerned with the ways in which discursive objects are
constructed. Which discursive object we focus on depends on our research
question. For example, if we are interested in how people talk about hove and
with what consequences, our discursive object would be love. The first stage of
analysis involves the identification of the different ways in which the
discursive object is constructed in the text. This requires that we highlight
are instances of reference to the discursive object.
- Discourses
Having
identified all sections of text that contribute to the construction of the
discursive object, we focus on the differences between constructions. What
appears to be one and the same discursive object can be constructed in very
different ways. The second stage of analysis aims to locate the various discursive constructions of the object within
side discourses. For example, within the context of an interview about her
experience of her husband’s prostate cancer, a woman may draw on a biomedical
discourse when she talks about the process of diagnosis and treatment, a
psychological discourse when she explains why she thinks her husband developed
the illness in the first place, and a romantic discourse when she describes he
she and her husband find the strength to fight the illness together.
- Action
orientation
The third stage
of analysis involves a closer examination of the discursive contexts within
which the different constructions of the object are being deployed. What is
gained from constructing the object in this particular way at this particular
point within the text. What is its function and how does it relate to other
constructions produced in the surrounding text. These questions are concerned
with what has been referred to as the action orientation of talk and text in
the previous chapter.
- Positioning
Having
identified the various constructions of the discursive object within the text,
and having located them within wider discourses, we now take a closer look at
the subject positions which they offer. A subject position within a discourse
identifies a location for persons within the structure of rights and duties for
those who use that repertoire.
- Practices
This stage is
concerned with the relationship between discourse and practice. It requires a
systematic exploration of the ways in which discursive constructions and the
subject positions contained within them open up or close down opportunities for
action.
- Subjectivity
This stage in
the analysis traces the consequences of taking up various subject positions for
the participants subjective experience. Having asked questions about what can
be said and done from within difference discourses. We are now concerned with
what can be felt, thought and experienced from within various subject
positions. for example it may be that positioning himself with in a discourse
of male sexual drive allows a man not only to publicity disclaim responsibility
for an act of sexual aggression, but to actually feel less guilty about it as
well.Psychology, The Science of Behavior
1
What is Science
Science or scientific methods
consists of the methods for collecting, analyzing, and drawing conclusions from
data.
Assumptions of all Science
Basics assumptions of all
sciences are the same. We will discuss these assumptions one by one.
Behavior is determined
Believe that there are specific
causes for each of our behavior is called determination. Near about all
psychologists believe on determinism, may be form is different. Like according
to Freud, childhood sexual traumas are the cause of psychopathology, in
contrast according to the behaviorism, faulty learning is the cause of
psychopathology. Some people reject the idea of determinism. They believe that
soul and mind control person’s every behavior, so there is no cause of any
behavior. This thought is referred to as will.
We can measure the critical variables
It can do the study of the
behaviors or event which can not seen or handling. I.e through using the study
of behavioral events we can draw inferences about mental events. Like scientist
use the different observations in order to draw inferences about the stars,
planets and other unapproachable objects.
What is Scientific Research?
Here we will discuss a number of
characteristics which make the research scientific.
Empirical Analysis
It means that researchers use
observations and experimentations for learning.
Public Verification
It means that if any one use the
same procedure (psychological instrument, research design, sample etc ), he or
she will get same results like us. This phenomenon is called public
verifications. Every one can verify the results after using the same procedure.
This is also called the repetition of the research or experiment.
Systematic Observation
It means that we observe, study
or compare the person’s behavior through using different condition in
systematic way. The goal of systematic
observations to examine the phenomenon under as many relevant situations as
possible. We continue to repeat the experiments to determine which conditions
produce the effect and determine whether other factors affect the phenomenon.
Control of Environment
In all form of research,
researchers attempt to control of environment in some way. We use this
phenomena to protect our research to be a confound. Researchers have the greatest level of
control when they conduct research in a laboratory setting because they can
control many or all of the environmental conditions. There are many conditions also,
when researchers are not able to control the environment, i.e. in naturalistic
observation or in field study etc.
Rational Explanation
It means a sound explanation that
behavior is determined (cause & effect) and follow a lawful pattern. Some
people give irrational explanations about cause and effect, this is called
pseudo explanation. Two common names of pseudo explanations are nominal fallacy
and the circular or tautological explanation.
The Foundation of Research
2
The Hypothesis in Research
Hypothesis
It is a prediction about the
relation among two or more variables based on a theory or on previous research.
Where Hypothesis come from
Hypothesis come from theory or
previous researches. We read previous researches and check out the extent and
direction of relationship between variables, then we draw our own hypothesis
according to them.
Theory
Theory a set of general
statements and claims that allow us to explain and predict various events. For
example. This is a theory that aggression and violence are learned behaviors.
In theory there will be an explanation of this statement that why people behave
violently.
Characteristics of Hypothesis
Hypothesis direct our Observation
When a researcher prepare a
hypothesis. He or she will focus their attentions on the type of he information
that they will study. Hypothesis identify the variables those the researcher
will examine.
Important
Terminologies related to Research
Variable
A variable is any characteristic
of a person, non human subject, environment, or research condition that can
have different values.
Constant
A numerical value that ,by
definition, can not change. Like 1 hour always contains 60 minutes.
Independent Variable
IV is the type of variable which
use to predict, describe, explain or getting change in dependent variable.
Dependent Variable
It is condition that the
researcher wants to explain, predict or describe.
Manipulated Independent variable
Variable that researchers
manipulate or control referred to as manipulated independent variables.
Subject variable
The subject variable is a
condition that the researcher can not manipulate or control. Like personality
types, age, sex etc.
Hypothesis Describe the Relation among Variable
One of the most important
component of the variable that it describe the relationship between variables.
Like depression is negatively related with self esteem.
Hypothesis refer to the Population
Hypothesis also gives us the
knowledge of population of our interest. Like if I prepared a hypothesis that “there
would be a positive correlation between self esteem and academic achievement
among adolescents” so here every one can under stand that I am going to conduct
a research on “adolescents” population.
Types of Hypothesis
There are four types of
hypotheses. Keep in mind that types of the hypothesis based on the type of the
prediction which the researcher want to make.
Estimation of Population Characteristics
The goal of this type of
hypothesis is to estimate the characteristics of a population using the information
from a sample. Researchers use few terms for estimating the population
characteristics like sample, descriptive statistics, parameters etc.
Correlation among two
Variables
Second type of hypothesis state
the correlation between two variables. Correlation means some thing link
together. It means if one thing increase, other either increase or decrease. When
we hypothesis about correlation, we can predict that there will be a positive
or negative or no correlation between the two variables. ) In positive
correlation two variables tend to move in the same direction. When the X
variable increases, the Y variable also increases and if the X variable
decreases, the Y variable also decreases. ) In a negative correlation. The two
variables tend to go in opposite directions. As the X variable increase, the Y
variable decreases. That is, it is an inverse relationship. Numerically the
signs of + or – indicate the direction of the relationship between variables.
The correlation will be in between the range of +1.00 to -1.00.
Difference among two
or more population
Some time we want to estimate the difference among two or
more population. He we use independent variable, dependent variable or subject
variable. And t-test and F-test etc will be use for statistical analysis.
Cause and Effect
Its very popular, famous and
demanding hypothesis type. In this hypothesis researchers want to predict or
explain that change in dependent variable is only due to the change in
independent variable. They want to discover the cause and effect relationship
between variables.
Measurement
Measurement is a very famous
concept in every type of science and research. Measurement allows the
researchers to record and quantify our observation in an objective and
consistent manner. Measurement is objective, when does correctly and researcher
can verify previous findings of specific phenomena.
Features of all forms of measurements
Operational Definition
Operational definition is a rule;
we use to convert our observation into numbers.
Scales of measurement
Nominal Scale: A nominal scale consists of a set of categories that
have different names. Measurement on nominal scale label and categorize
observation, but do not make any quantitative distinctions between
observations. Although the categories on a nominal scale are not quantitative
values, they are occasionally represented by numbers. For examples room and
offices in the building may be represented with numbers. Like room no 1, room
no 2, room no 3 and so on.
Ordinal Scale: An ordinal scale consists of a set of categories
that are organize in an ordered sequence. Measurement on an ordinal scale rank
observations in terms of size or magnitude. Like! First position, second
position and third position, intense fear, low fear and minimal fear etc.
Interval & Ratio Scale: An interval scale consists of ordered
categories where all of the categories are intervals of exactly the same size.
With an interval scale, equal differences between numbers on the scale reflect
equal differences in magnitude and ratio scale having the additional feature of
absolute zero. For example! Zero is meaningful, 0 height means no height or
zero weight means no weight.
Reliability of Measurement
Reliability refers to the
consistency of measurement. A reliable test will produce similar result each
time we measure the same thing. Unfortunately no test available which is
perfectly reliable because measurement error effect all measurement techniques.
The classical theory or test reliability states that any test score is the sum
of the true score plus the effects of measurement error.
Test score = True
score ± Measurement Error
The symbol in the equation
indicates that measurement error can add to and remove from the true score of
the construct that we want to measure. One of the common methods to improve the
reliability of the data is to increase the number of measurement. Like good
carpenters always measure the things twice or thrice before cutting or
drilling.
Validity of Measurement
Validity refers to degree to
which a test or research method accurately measures the construct of interest. If
we say that test is not valid, we imply that the test measures things unrelated
to the construct we want to measure.
Types of Validity
Face Validity
Appearance of test is matter. If
a test is not authentic by face, it will be not valid. If the people taking the
test do not believe that the procedure is valid, they may not respond honestly.
Content Validity
Means that content of the
measurement tool should be the relevant to the targeted behavior (which
researcher wants to measure). For example if I am going to measure the
depression level of some one, definitely I will use the questionnaire; comprise
of the depression related questions.
Predictive Validity
Any good and valid test allow us
to predictive the other behaviors. For example test used in entry test predict
the student’s behavior that, what type of performance they show after admission
in any institution.
Concurrent Validity
Concurrent validity means test scores
should be relevant to the other similar behaviors. i.e if researcher administered
the self esteem scale on any person, then scores should be similar to the
scores find from self concept scale of similar person.
Survey of Empirical Methods
3
Internal and External Validity
Internal validity refers to the
cause and effect relation between the independent variable and the dependent
variable. An experiment is internally valid means that there is no doubt that
changes in the independent variable, which the experimenter controlled, caused
changes in the dependent variable.
Threats to Internal Validity
Any threat to internal validity
is a confounding variable. Confounding variable is an uncontrolled condition
related to the independent variable. Researcher can not control or manipulate
the confounding variable. There are four main categories of threats to internal
validity. Let me explain one by one.
Unintended sequence of events
The first threat to internal
validity is the unintended sequence of events. As the name implies, the data
may reflect a sequence of events that the researcher did not or could not
control.
Carryover effects
When one part of the study effect
the other part of the study. Or when researcher administered the same test for
two time. Participant’s experience after participation first time in the test
help him in second time. Which is major threat to the internal validity of a
research.
Maturation
If the research project covers
the long time. It is possible that the change in DV is due the long time and
person’s maturity. For example some depressed patients gradually improved with
our taking any treatment or therapy.
Intervening event
If the research covers long time,
participant may be experience any event which become the cause of the change in
dependent variable. For example during study participant lost his loved one,
which will draw an adverse effect on his mood, lead to adverse effects on
results of the study.
Non equivalent Groups
The second threat to internal
validity is the problem created by non equivalent groups. When we want to study
the effects of independent variable on dependent variable. The both groups
should be equal except independent variable. In the case of non equivalent groups, research could not be
able to draw the conclusion that either change in DV is due to the IV or the
other confounding variable. Like age, gender or sex etc.
Mortality
Mortality does not solely means
the death of the participants, it refers to the termination of the participants
from research study during continuation of the study. It may be due to the participant’s death, bored ness or migration
from one city to an other city or country.
Subject Variable
For many research projects, the
independent variable is a subject variable. Because the researcher does not
have direct control over this variable, he or she can examine the relation
between the independent and depend variable, but can not automatically assume
cause and effect.
Measurement Error
This threat is related to the
measuring instrument, which is not valid for measurement of any construct.
Ceiling and Floor effects
Some time researcher may select
very easy test, which attempt every one, or some time very difficult test,
which is not in the approach of any one. In this way researcher will be not
able to find the differences among the participant’s score.
Reliability
Due to the invalid test,
reliability of the score will be also not sufficient.
Regression to the Mean.
Change in score represent the regression
to the mean phenomena. This effect occurs when people receive exceptionally
high or low scores on a test or measure. When tested a second time, the scores
will end to be close to the average. For example some people may receive a poor
score on a test because they do not feel well or for reasons unrelated to their
knowledge of the material. When we test them again a few weeks later, their
scores will probably be close to normal consequently, the change in test scores
may reflect regression to the mean and not the effects of the independent
variable.
Ambiguity of cause and effect
The final threat to internal
validity refers to the assumptions of cause and effect. We must show the cause
occur before the effect.
External Validity
External validity refers to the
generalization drawn from the data. it is related to the representative sample.
Sample will be more representative, research will be more externally valid.
There are two types of external validity. 1) Generality of finding refers to
generalization from sample to population. 2) Generality of conclusion refers to
the generalization from population to population.
Threats to External validity
There are three major threats to
external validity for any research. 1) recruitment of participants, 2) effects
of situation, and 3) effects of history. Let me elaborate these.
Recruitment of Participant
What is our method of selecting
participants for our research project. Either we are going to use random
sampling or convenient. Some time our method of recruitment of sample does not
allow us to make generalization of our results on large population. For example
we are going to find achievement level of adolescents, and we are going to draw
sample from university, is it possible that university adolescents sample will
represent the entire population of adolescence who work in shops, hotels or
related to the other type of occupations.
Effects of Situation
Situation is also matter in
research. We can not generalize the findings of one situation to another situation.
For example I can not generalize the findings of my university to the Oxford University
Effects of History
Can the results of research
conducting before 50 years generalize today? Time and situations matter in this
long history. Before 50 years female was not more efficient in mathematics than
today. Today they are provided more facilities to learn math.
Survey of Empirical Methods
In this section of the notes, we
will discuss some important methods of conducting research in science,
specially in social sciences. Let me explain gradually.
True Experiments
This method use for determining
the cause and effect. True experiment is very useful due to his quality of
control and remove the threats to internal validity, which are very important
for any successful and good research. It ensure that IV comes first and DV
comes later and it is also ensure that IV and DV are correlated. In this method
researcher can manipulate and control the IV and researcher use random assignment
for selecting the participants to the conditions.
Random Assignment
Random assignment means that each
participant selected for the experiment as an equal chance of experiencing the
one of the research condition.
Use of Control group
A control is the group of
participants who experience the same treatment as the other participants except
for exposure to the independent variable.
Intact Group Design
This method use when researcher
can not manipulate or control IV and use subject variable. In this method
researcher does not select the sample from a single population, several
population use for sampling which represent their subject variable. For example
patients of different disorders. Here we can not use random assignment due to
taking subject variable. Here we collect data and compare the difference among
groups.
Quasi Experimental Design
A quasi experiment is similar to
a true experiment except that we can not randomly assign the participant to the
participant to the treatment and control condition. The advantage of this
design to compare the groups. Like male vs. female etc. limitation of this
method that it can not explore the cause and effect. Here third variable
problem is also important. For example we don’t know that change occur due to
IV or any other variable.
Survey
Researchers use survey to
estimate the population parameter. It uses of find the opinions and attitude of
the population toward a specific concept, problem, issue or phenomena. The
advantages of the survey are ! it is not expensive, done quickly and provide a
large number of data from the population.
Correlational Studies
Correlation is used to measure
and describe a relationship between two variables. Usually the two variables
are simply observed as they exist naturally in the environment. There is no
attempt to control or manipulate the variables. Correlation can be classified
into two basic categories. Positive and negative correlation. 1) In positive
correlation two variables tend to move in the same direction. When the X
variable increases, the Y variable also increases and if the X variable
decreases, the Y variable also decreases. 2) In a negative correlation. The two
variables tend to go in opposite directions. As the X variable increase, the Y
variable decreases. That is, it is an inverse relationship. Numerically the
signs of + or – indicate the direction of the relationship between variables.
Range of the relationship between two variables will be between +1.00 to -1.00.
This is called coefficient of
correlation”. Co relational studies help the researchers to better understand
the complex human behaviors and again it is not able to find out the cause and
effect between variables.
Single Participant Research Methods
Single Participant experiments
This method use one participant.
Mostly use in experimental researches in order to determine the IV and DV
relationship and determine the cause and effect between variables.
Case Study
This is one of the most familiar methods
of single participant research method. In this method researchers study a
single participant in deep. The essential feature of this method is intense
review of the behavior of a person. The advantage of this method is that we can
use sound experimental techniques to determine how to influence a person’s
behavior. The disadvantage of this method is that researchers can not
generalize the results of a single person to the whole population.
A collection of statistical
techniques used to combine the results of separate research projects to help
one determine the relation between the independent and dependent variables. This
method used widely in psychology, sociology, political science, medicines and
other researches.
Learning About Psychology and Forming Hypotheses
4
Bibliographic Research
Three are three important levels
of bibliographic sources. We will discuss these sources one by one.
Tertiary Bibliographic Sources
This is third level source.
Textbook, articles, news papers, psychology dictionaries, encyclopedias are all
tertiary sources. Author of these sources did not review the topics in detail.
The advantage of a text book is that it gives you a quick and easy to
understand introduction to a topic. This advantage is also limitation.
Secondary Bibliographic Sources
These sources more emphasis on
the topic than tertiary sources. These are the comprehensive reviews and author
discuss on the topic in details. Text books having many chapters on a single
topic, discuss its current trend and theories are also include in secondary
sources. It also give the detail of method of conducting research on the
specific phenomena. This source emphasis on different issues related to the
research topic.
Literature Review
This is not secondary source but
this shares some features with secondary source. Literature review found in
professional scientific journals. Literature Reviews are more topic specific
than text books. It is also provide a comprehensive summary of current
research. Different journals are the best sources of getting literature review.
Primary Bibliographic Sources
Original research reports, you
can find in different journals are considered as primary bibliographic sources
of research. The advantage of the primary source is that there is no filtering
of information. These research reports often difficult for the readers, authors
of research reports of ten use these reports to communicate their research
conclusion to the other researchers.
Peer Review
This is the very good
characteristic of research journals and text books. The goal of the peer review
is to filter out bad research and make the good research as clear and
compelling as possible. Peer review is a part of the normal edition process
during which experts read a manuscript to ensure that the researchers used
appropriate methods to collect and analyze the data and then made reasonable
inferences from the data. when a researcher submit his or her research report
to the management of any journal, a group of the research mostly three to four researches,
having enough knowledge on the relevant topic discuss and review the research
report. After reviewing either may be the accepted, sending the report to the
author for amendment or rejected.
Developing a Search Strategy
Searching the Literature: The Library
A library is a wonderful place to
find information. How you can use your library. We will help you. Let me
explain !
Library Staff
It’s a very common thing, that
student ignore the very important aspect of searching from the library, it is
library staff. Librarians have the knowledge about the storage of books that a
book kept in which cupboard? Almost staff of every library know very well this
type of information.
Search Tool
Card catalogue and different
reader’s guides are the helpful for searching any book, journal or any other
source of searching literature.
Online Catalog
All libraries keep their
bibliographic material in a well organized system that you can use to find
information quickly. Specifically, most colleges use the library congress
classification scheme that uses a combination of letters and numbers to classify
each book, document, other materials such as films and records.
Sampling, The First Steps in Research
6
The nature of Sample
A number of ways that define
samples. First, sample is a valid representation of a population. Second way is
that compare the sample with those things, those are not sample. A true, valid
and successful research depends on the nature or type of sample. Type of sample
depends on that what type of method is used for data collection. Now we discuss
the type of the sample.
Scientific samples versus Anecdotal Evidence
Anecdotal evidence means
collection of personal experiences haphazardly that may be biased. For example
I went to Lahore ,
stay there just for 2 hours and came back. Here I draw inference that people of
Lahore are very
cruel. Is it true ? no its wrong, because this inference is depend on my
personal experience. May be I will met there just 10 to 15 people, those are
not the true representative of 500000 people of Lahore . And a scientific sample means that
the sample select by using true and good sampling technique and sample should
be the true representative of the population.
Probability Sampling
Probability sampling means that every one from the
population has an equal change of being selected for a research study.
Some important terms related to sampling
Sampling population
Selected group of people, who share the same characteristics
as the target population.
Subject Pool
The subject pool refers to a group of people to whom the
researcher has an easy access.
Sampling Frame
A list of names that identities the members of the
population is called sampling frame.
Characteristics of Probability Sampling
- Treat each member of the population as a potential member of a sample.
- We can determine the probability of selecting any specific sample. For example you have a population of 6 participants and you choose 2 of them. You will draw 15 potential samples (pair of two people). For example you have 1, 2, 3, 4, 5, 6. how much pair is possible (1, 2) (1, 3) (1, 4) (1, 5) (1, 6) (2, 3) (2, 4) (2, 5) (2, 6) (3, 4) (3, 5) (3, 6) (4, 5) (4, 6) (5, 6). And now you can randomly select any pair (sample of two participants).
Sampling Methods
There are many methods of drawing
samples from the population. Each method shares a common goal, ensure that the
sample is an unbiased depiction of the population.
Simple Random Sampling
Simple random sampling is occur
when each member of the population has an equal probability of selection. In
this method first researcher estimate the size of population then determine the
random number for selecting the participant in research.
Sequential Sampling
In sequential sampling,
researcher list the member of the sampling population and then select each kth
member from the list by using interval. Interval can be determine by using this
method
Sampling Interval =
Size of population ⁄ desired size of
sample
If researcher got 3 answer, then
it means that every 3rd person will be selected for sample.
Stratified Sampling
This technique is especially
useful in situations where the population contains different subgroups that the
sample must include. For example age, education, socioeconomic status etc. for
selecting the sample, in initial step researcher identify the subgroups and
then use simple random sampling or sequential sampling in subgroups for draw
sample.
Cluster sampling
This method use when it is
impossible to take one or two people out of their environment or when other
sampling methods are not possible or impractical. Researchers often use this
technique when working with intact groups. For example researcher is interested to find
the impact of teaching style on the students. There are 20 different sections
of the same subject of psychology. Here researcher use cluster sampling
technique.
Non Probability Sampling
In many situations researcher
will resort to non probability sampling. Using a non probability sampling
method may limit he inferences.
Convenience Sampling
Convenient sampling means that
the researcher uses members of the population who are easy to find. Interviewing
people in shopping malls are the example of convenient sampling.
Snow ball sampling
Some time it is not possible or
difficult for the researcher to find the members of a coherent group for the
research. In such situations, the researcher may wish to use snow ball
sampling. The researcher needs to find a member of the cohort and use him or
her to find other member of the cohort.
CREATING AND USING TESTS, SURVEYS, AND OBJECTIVE MEASURES
7
Measurement is the heart of
empirical sciences. There are two goals of measurement 1) for replacing the
ambiguity of words with operationally defined constructs. For example
operational definition of depression may be lack of sleep, lack of sexual
activity, sadness ete. 2) Second goal is standardization and consistency of
measurement. Consistency means compare people using a common set of procedures
and scales. Standardization means that number have constant meaning for every
one for every culture. Like 110 will be same in Pakistan , India , Germany and
Switzerland etc.
There are two common
misinterpretations of psychological tests and measurement.
Human behavior can’t be measure (
human behaviors are too complex to measure objectively. This is common complain
because some people think that Psychology is not a science.
Uncritical trust on numbers. This
problem is called mystique of quantity which means an exaggerated regard for
the significance of measurement, just because it is quantitative with out
regard to what has been measures (behaviors).
Creating a Measurement Scale
What questions are you trying to answer ?
Best way to answer this question
is to read the hypothesis. Well formulated hypothesis make you able to define IV and DVs. For
example if a researcher formulated the hypothesis that Depression is negatively
correlated with self esteem. It means that researcher want to calculate the
relationship between depression or self esteem (either increase or decrease in
depression draw effect on self esteem). Identifying the variables is half the
battle as it helps you identify the types of behaviors you want to assess.
What is the most convenient method for producing data
It depends on the nature of iv
and dv. For example if researcher want to check out the children’s behavior of
different ethnic groups, you will have to use observational technique for data
collection but if you want to compute the relationship between self esteem and
depression of the participants, you will use questionnaire for data collection.
What is most accurate measurement technique?
There is no perfect measurement
technique because all techniques have some sort of errors. Well you should use
methods which reduce the bias. For example for observational method you should
use some trained observers who observe and save all necessary and relevant
samples of behaviors which is needed.
Constructing Interviews, Questionnaires and Attitude Surveys
There are many ways that we can ask people questions like
face to face interviews to questionnaires. We will discuss one by one.
Personal Interviews:
Personal interviews are either
face to face or telephonic. It is very popular technique for gathering data
regarding person’s attitudes and beliefs. Specially face to face interviews
have high level of cooperation. Interviews can ensure that participants
understand the questions. Interviews can highly structures or unstructured may
be individual or may be in group setting. Along with advantages interviews have
also some disadvantages. For example its costly and time consuming, conducting
interview require much practice. Conducting interviews difficult when there are
cultural, language and racial barriers between interviewers and participants.
Self Administered
Surveys
Alternative of personal interview
is self administered surveys. Give many people questionnaires and receive back.
Advantages are: it is not costly, no required any special training, not time
consuming. (Use a captive audience) Use
available participants for getting information through surveys. For example
present students in introductory psychology class. It is very quick and easy
approach method. (Use social psychology
to your advantage) You can also use social psychology methods for
completion your surveys. It gives you the methods that how you can collect most
filled questionnaires form the participants.
Preparing the questions
When you select the method for
obtaining data, you have needed to prepare the questions. You can ask two types
of question 1) open ended 2) close ended.
Open ended questions (OEQ)
It gives the opportunity to
person to give short answer to the question.
For example
Do you satisfy with your
statistics teacher? In the answer of this question respondent can give the
short answer
Closed response questions (CRQ)
In CRQ researcher supply response
option to the participants. Researchers use many alternative formats for CRQ.
Nominal Category Response
I am: Single Married Divorced Widowed
Forced Choice Response
Participants asked to select the
response that best represents their answer to the question. For example
Shy vs Out
going
Leader vs Follower
Introverted vs Extroverted
Likert Format
It is one of the most popular CRQ
format. it offers a clear and unambiguous ordinal scale of measurement. You can
use same format for many questions.
Likert scale format consist of 4 to 10 points that represent potential reaction to the question or
statement. There will be a middle neutral option in the likert scale format.
for example
Overall, I believe that Chief Minister Shehbaz Sharif is doing a good
job
Strongly agree Agree Neutral Disagree
Strongly disagree
Guttman Format
A guttman format like the likert
format, represents a continuum that ranges from one extreme to another.
For this format researcher arrange the answer options by levels of acceptance. For
example
Which of the following best
represents your opinion regarding abortion? Select only one statement
- Abortion may be performed when the mother’s life is in danger.
- Abortion may be performed if the fetus has a severe disability
- Abortion may be performed in case of rape
- Abortion may be performed as a type of family planning
- Abortion may be performed for any reason.
Writing Good Questionnaire and Survey Items
Ask Single questions
or make single statement
Example: Do you agree with the college’s plan to build a new
student recreation center and increase the student fee to 500 dollars to help
pay for the building ?
Alternative: Do you agree with the college’s plan to build a
new students recreation center ?
Ask Specific
Questions and Avoid vague terms
Example: Do you actively support the college’s decision to
build a new student recreation center.
Alternative: I believe that the college should build a new
student recreation center as soon as possible
Write Neutral
Statement and Avoid a Biased Tone
Example: Many people believe that it is bad to spank children.
Do you ?
Alternative: Do you believe that parents should spank their
children.
Ask questions that
Don’t embarrass or Anger the Participant
Example: Are you still a virgin?
Alternative: With how many people have you had sexual intercourse?
Whenever possible,
Use Simple Words and Concepts
Example: Do you
believe that you received sufficient time to consider your alternative before
you were required to purchase your computer?
Alternative: Did you feel rushed to buy the computer?
Recognize one question
may not be enough
Example: All in all, I am inclined to feel that I am a
failure?
Alternative: I feel that I have a number of good qualities.
I am that I have a number of good
qualities.
I feel that I do not have much to be
proud of.
RESEARCH DESIGN I:
9
(Between Subject Design)
A research design refers to the methods used to collect data
that will decisively answer an empirical question. A good research design must
have three characteristics. 1) It must be efficient (design should produce the
best quality data in least amount of time and money). 2) The design must
eliminate factors those biased the results. 3) Design should produce the data
which clearly address the research question.
A MODEL FOR RESEARCH DESIGN
The purpose of any research
design is to protect researchers from lapses in objectivity and the bias inherent in
personal judgment. Especially in Psychology the participants come to us
with their anxieties, fears, biases and awareness that they are coming for
participating in research study. Good behavioral scientists will be those who
collect very useful data from those participants.
Purpose of Research
Here the first thing is the
personal bias of the researcher. Mostly research questions based on
researchers own experience. Every one explains the phenomena according to their
own self. The influence of researcher’s orientation can have considerable
influence on how he or she will view a problem.
Design of Research (control for extraneous variables)
After identifying the purpose and
the hypothesis, the next step is the design phase during which the researcher
develops controls for potential confounding variables. Researcher
formulates a design that what are the extraneous variables and how he or she
can control them during study.
Selection of Participants
This is the stage of selecting
the representative sample for the research. Most of the problems and the
threats to the internal validity occur due to the non representative samples,
which do not produce the true data for making inferences for the population.
Preparing the Participant for the Study
Independent Variable
Variable which use the researcher
for making prediction about independent variable?
Type of Independent variable
Type of independent variable is
draw a sound effect on the design of the research. It is also important in the
context of control conditions. There are
two type of IV (subject vs. manipulated). subject variable is the variable
which exist before the participation of the person in the study, like age, sex,
personality traits etc. manipulated variables are the variables which
researcher manipulate on the participants, like researcher select the alcohol
quantity for administering on the participants.
Between Subject versus within subject variables
We can also make a comparison of
IV in the context of between subjects or within subject variables. In between subject variable, participants
tested only under in one treatment or condition and in within subject variable,
participants tested under more than one treatment conditions. When we compare
the depression level among genders then it means that gender (male &
female) are the between subject variable because male type of sex did not give
the correct estimate of depression level of female gender and vise versa so the
sex is representative of between subject variable. On the other hand when researcher
observes the behavior of the same person under different treatment conditions
or level of IV so it is representative of with in subject variable. For example
when researcher observe the behavior of a person under different time table.
Dependent Variable
Dependent variable is the variable
which changes due to the change in independent variable. Researcher can
measure the dependent variable by evaluating the construct of dependent
variable. Like if researcher is interested in computing the effect of self
esteem on depression so in this case depression is an independent variable and
researcher will measure the depression by evaluating its contents for example
lack of sleep, lack of hunger, low interest in sexual activities etc.
Reliability and Validity
Reliability means the consistency
of scores and validity means the use of true instrument for measuring the
construct. Low reliability and validity of research projects is due to the use
of techniques that were unreliable, invalid or both. There are many ways to
ensure the reliability and validity of a measure. One is to use techniques that
other researchers use.
Objective Recording of Data
One important and useful method
for reducing experimenter bias single blind or double blind methods. The single
blind procedure indicates that the person participating in the research does
not know the specifies of the researcher’s hypothesis. The goal of using single
blind procedure is to reduce the effect of demand characteristics of the
participants. Demand characteristics means that participants having the
knowledge of the participants hypothesis and then participants behave according
to the hypothesis. In double blind studies participants and as well as
researchers did not know about the
research objectives.
Placebo
Placebo is a common technique
used to create a single blind. For example when researcher want to examine the
effect of a drug, he can use placebo along with real drug. Participants of the
research did not know that either they receive real drug or fake pill
(placebo). Placebo is also use in control group.
Carry over effect
Carry over effect refers to the
effect that previous experience have on subsequent behavior. As an example, if
researcher uses the same test on two occasions, the participants may remember
the questions and their answers, and this memory may bias the results.
Research Hypothesis
Hypothesis is the tentative
explanation of prediction of the relationship between two variables. Preparing
hypothesis is very important part of any research design. Hypothesis not only shape the
design of the research but effect on statistical analysis of the data.
Hypothesis tells the relationship between the groups. This relationship may be
directional or non directional. In this context research divide the hypothesis
into two types. 1) Directional hypothesis 2) Non directional hypothesis.
In directional hypothesis
researcher indicate the relationship between groups in the form of greater than
or less than. For example Women are more competent in mathematics as compare to
men. In non directional hypothesis researcher indicate the relationship between
groups but no direction of the relationship. For example there would be a
relationship between depression and self esteem among college students.
Mathematical Hypothesis
When researcher use the
mathematical impression in writhing the hypothesis, this form of
hypothesis is called mathematical hypothesis. When researcher uses the
mathematical impression for of hypothesis, they use two types of hypothesis 1)
first is Null hypothesis and second is Alternative Hypothesis. Alternative
hypothesis is also called research hypothesis or scientific hypothesis. Null
hypothesis indicates that there would be a no difference between populations
and alternative or research hypothesis means that there would be a difference
between groups.
Evaluating Hypothesis
When researcher use data for make
prediction, they use the term rejecting or fail to reject of null hypothesis.
When researcher make inference, they keep in mind the probability of doing
mistakes in rejecting or fail to reject the null hypothesis. There are two
types of mistakes or errors. 1) Type 1 error 2) Type 2 error. Type 1 error is
that when researcher reject the null hypothesis when it is true. This is due to
the mislead of data. It is due to the extreme samples and other random events.
For example difference between the groups. Type 2 error is when researcher is
fail to reject the null hypothesis when it is truly wrong or rejected.
Research Design II
Single Variable Between-Subjects Research Design
This research design is related
to single independent variable with more than two conditions. For this design
researchers use ANOVA .ANOVA allows us to examine the effect of many different
levels of the same independent variable on the dependent variable.
Research Design III
Between-Subject Factorial Designs
This research design incorporate
more than one independent variables. This design has the following advantages
1) examine the effect of more than one independent variable 2) examine the
interaction between independent variables. Two variable design examines how
much each independent variable, by itself, influences the dependent variable.
In addition ,the test determines how much the combination of the independent
variables affects the dependent variable.
2×2
It means that two
variables and two two conditions in each variables
Main Effects
A main effect refers to the effect that one independent
variable has on the dependent variable holding the effects to the other
variable constant.
Research Design IV
Correlated-Groups Designs
Repeated measure designs or
matched groups designs are called correlated groups designs. These designs
allow the researcher to exercise greater control over the variables that
influence the dependent variable. The main characteristic of this design
distinguished it from other designs is that it has some known or intentional
correlation among the variables. In this design researcher intentionally create
the groups in such a way that there is a correlation among pairs of scores
across the groups. Two primary techniques that researcher use is repeated
measure design or matched group design. Both are different to each other.
The logic of using these
techniques is to minimize the individual differences among participants which
play a very vital role in within group variance. There are several ways we can
reduce the within groups variance like
selection of more homogenous sample, second is to treat more significant
subject variable sex as one of the independent variable in a factorial design.
Another way to reduce the within groups variance is to use the correlated
groups design. This procedure is an exceptionally cost effective method for
increasing statistical power,. Like the factorial design, the correlated groups
design allows the researcher to reduce the within groups variance.
Repeated Measure Design
In the repeated measures design,
we collect the data from the same participants on a series of occasions or
under different levels of the independent variable to determine if there are
systematic changes in the participant’s behavior. There are two types of
repeated measure design.
The first type of repeated
measure design allows us to test the same participant using several levels of
the same IV. The advantage to this procedure is that it requires fewer
participants because we test each participant using each treatment rather than
assigning separate groups. Another advantage to this design is that we can use
each participant as his or her own control condition and thereby reduce the
size of the within groups variance in the ANOVA. The second type of repeated
measure design allows us toe use time as an independent variable. For this type
of research, we arrange to test or sample the participant’s behavior changes as
a function of time. This technique is useful when time is a critical predictor
variable.
Grounded Theory
Basic Principles of Grounded Theory
Building Blocks
Grounded theory as method
provides us with guidelines on how to identify categories, how to make links
between categories and how to establish relationships between them. Grounded
theory as theory is the end product of this process: it provides us with an explanatory framework
with which to understand the phenomenon under investigation.
Categories
These designate the grouping
together of instances for example events etc. that share central features of
characteristics with one another. For example anxiety, anger and pity under the
heading of emotions.
Coding
This is the process by which
categories are identified. In the early stages analysis, coding is largely
descriptive. Here descriptive labels are attached to discrete instances of
phenomena.
Constant comparative analysis
This ensures that the coding
process maintains its momentum by moving back and forth between the
identification of similarities among and differences between emerging
categories. Having identified and common feature that unites instances of p
phenomenon, the researcher needs to refocus on differences within a category in
order to be able to identify any emerging subcategories.
Negative Case analysis
This ensures that the researcher
continues to develop the emerging theory in the light of the evidence. Having
identified a category, or a linkage between categories, grounded theory
researcher need to look for negative cases that is, instances that do not fit.
The identification of such instances allows the researcher to qualify and
elaborate the emerging theory, adding depth and density to it, so that it is
able to capture the full complexity of the data on which it is base.
Theoretical sensitivity
This is what moves the researcher
from a descriptive to an analytic level. In grounded theory, the researcher
interacts with the data. That is , he or she asks questions of the data, which
are in turn modified by the emerging answers. Each emerging category, idea,
concept or linkage informs a new look at the data to elaborate or modify the
original construct
Theoretical sampling
This involves collecting further
data in the light of categories that have emerged from earlier stages of data
analysis. Theoretical sampling means checking emerging theory against reality
by sampling incidents that may challenge or elaborate its developing claims.
While the earlier stage of grounded theory require maximum openness and
flexibility to identify a wide range of predominantly.
Theoretical saturation
Ideally, the process of data
collection and data analysis in grounded theory continues until theoretical
saturation has been achieved. In other words, the researcher continues to
sample and code data until no new categories can be identified, and until new
instances of variation for existing categories have ceased to emerge.
Memo writing
This is an important part of
grounded theory. Throughout the process of data collection and analysis, the
researcher maintains a written record of theory development. This means writing
definitions of categories and justifying labels chosen for them, tracing their
emergent relationships with one another, and keeping a record of the
progressive integration of higher and lower level categories. Memos will also
show up changes of direction in the analytic process and emerging perspective,
as well as provide reflections on the adequacy of the research question.
Memory Work
Data Collection and analysis
This method involves a process of
group work during which co researcher remember, compare, discuss and theorize.
Memory work research can take many months or even years to complete, and it can
generate diverse insight.
Phase: 1 Generating memories
Step 1: Form a memory
work group
Memory work groups can have
between four and eight members. Such group share characteristics which may be
regarded as relevant in relation to the topic under investigation. For example
with topics such as emotions or sexual practices., single sex groups may be
preferable. For some topics, a group of friends may be best, whereas other
issues may be best discussed within an atmosphere of relative anonymity. The
most important requirement, is mutual trust and a feeling of safety within the
group.
Step 2 Choose a trigger
To prompt group members to write
memories, the group needs to agree on a suitable trigger. The trigger is a word
or a short phrase which is expected to generate memories that are relevant to
the topic under investigation.
Step 3 Write memories
Each member of the group writes a
memory in response to the trigger. The writing of memories can take place
either within the context of the group meeting itself or outside of it. To ensure
that the memories are written in a way which renders them suitable for memory
work analysis, they should be written in the third person singular and include
as much circumstantial detail as possible. This could be references to sounds,
tastes and smells that appear to be inconsequential ofr trivial in terms of the
event being remembered. It is important to avoid censoring apparently
irrelevant information and to avoid interpretation or justification. The style
of writing should be purely descriptive biographical.
Phase 2: Analysis of the memories
Step 4: Texture analysis
The memory work grouper convenes,
with members bringing along their written memories. Initially, each individual
memory is analyzed separately. Each member of the group is handed a typed copy
of each memory that has been produced. The memories are then examined, one by
one, in terms of the sequences of actions, role relations, clichés and contradictions,
statements made and absences which characterized them.
Step 5 Cross sectional analysis
Here the co researcher compare
their memories with one another. They look for similarities, differences,
recurring themes and common patterns.
Step 6 Re writing the memories
At this point in the process of
analysis, it may be helpful for group members to return to their memories and
to re write them in the light of the textual and cross sectional analysis. Re
writing memories can be a way of becoming more acutely aware of the pro esses
of meaning construction because it requires us to thing of different ways of
making sense of the same event. This is where memory work connects with
consciousness raising.
Phase 3: Integration and theory building
Step 7: Analysis of transcripts and memories
Now it is time to compare and
contrast the memories produced in response to several trigger. The transcripts
of the group discussions of the memories constitute further data for integrative
analysis. The ideas generated by the group in relation ot particular social
constructions are themselves subjected to critical appraisal and further
theorizing.
Step 8: Writing memory work
The process of writing up memory
work is still part of phase 3. writing both requires and facilitates
integration of material it is also likely that theoretical formulations will be
clarified, and even revised, during the process
of writing. Since memory work is a collective enterprise, writing can
not be separated from further discussion. Drafts of papers will have to be read
and discussed by the co researchers, and this may generate new ideas and
changes to the original analysis.
Interpretative Phenomenological Analysis
IPA works with transcripts of
semi structured interviews. Smith provides guidance on how to conduct
interviews hat will generate data suitable for phenomenological analysis. In
general, interviewing for IPA shares the principles and practices associated
with semi structured interviewing. Since phenomenological research requires the
researcher to enter the life world of the research participant, it is extremely
important that the questions posed to the participant are open ended and non
directive. Their sole purpose is to provide participants with an opportunity to
share their personal experience of the phenomenon under investigation with the
researcher. Focused and specific questions should be used toe encourage participant
to elaborate rather than to check whether they agree or disagree with particular
claims or statements. Diaries, video or written means also use for the data
collection for IPA. IPA use ideographic approach. One this is more important in
the IPA, that researcher’s point of view about the phenomenon is also include
in the interpretation.
The first stage
of analysis in IPA involves reading and re-reading of the text. Researcher produce wide ranging
notes that reflect the initial thoughts and observations he or she may wish to
record in response to the text. These could include association, questions,
summary statement and open coding like grounded theory.
- Identification
of themes (each section)
The second stage
of analysis requires the researcher to identify and label themes that
characterize each section of the text. Themes titles are conceptual and they
should capture something about the essential quality of what is represented by
the text. Psychological terminology may be used at this stage.
- Clustering the
themes
The researcher
listed the themes identified in stage two and thinks about them in relations to
one another. Some of the themes will form natural clusters of concepts that
share meanings or references, whereas others will be characterized by
hierarchical relationships with one another. For example themes such as
childhood memories, going to school and relationship with mother could form a
childhood cluster.
- Production of
a summary table
The forth stage
of analysis production the summary table of structured themes, together with quotations
that illustrate each theme. The summary table should only include those themes
that capture something about the quality of the participant’s experience of the
phenomenon under investigation. This means that some of the themes formulated
in the stage 2 will be not included in the summary.
Discourse Analysis
- Discursive constructions
The first stage
of analysis is concerned with the ways in which discursive objects are
constructed. Which discursive object we focus on depends on our research
question. For example, if we are interested in how people talk about hove and
with what consequences, our discursive object would be love. The first stage of
analysis involves the identification of the different ways in which the
discursive object is constructed in the text. This requires that we highlight
are instances of reference to the discursive object.
- Discourses
Having
identified all sections of text that contribute to the construction of the
discursive object, we focus on the differences between constructions. What
appears to be one and the same discursive object can be constructed in very
different ways. The second stage of analysis aims to locate the various discursive constructions of the object within
side discourses. For example, within the context of an interview about her
experience of her husband’s prostate cancer, a woman may draw on a biomedical
discourse when she talks about the process of diagnosis and treatment, a
psychological discourse when she explains why she thinks her husband developed
the illness in the first place, and a romantic discourse when she describes he
she and her husband find the strength to fight the illness together.
- Action
orientation
The third stage
of analysis involves a closer examination of the discursive contexts within
which the different constructions of the object are being deployed. What is
gained from constructing the object in this particular way at this particular
point within the text. What is its function and how does it relate to other
constructions produced in the surrounding text. These questions are concerned
with what has been referred to as the action orientation of talk and text in
the previous chapter.
- Positioning
Having
identified the various constructions of the discursive object within the text,
and having located them within wider discourses, we now take a closer look at
the subject positions which they offer. A subject position within a discourse
identifies a location for persons within the structure of rights and duties for
those who use that repertoire.
- Practices
This stage is
concerned with the relationship between discourse and practice. It requires a
systematic exploration of the ways in which discursive constructions and the
subject positions contained within them open up or close down opportunities for
action.
- Subjectivity
This stage in
the analysis traces the consequences of taking up various subject positions for
the participants subjective experience. Having asked questions about what can
be said and done from within difference discourses. We are now concerned with
what can be felt, thought and experienced from within various subject
positions. for example it may be that positioning himself with in a discourse
of male sexual drive allows a man not only to publicity disclaim responsibility
for an act of sexual aggression, but to actually feel less guilty about it as
well.
No comments:
Post a Comment