MPC-005 Research Methods

Section A (1000 words)

Define sampling. Discuss methods (Block 1, Unit 4)

Sampling is a key technique in any research. It turns the problem of collecting data from a large and
widely dispersed set of individuals (population) into one of accessing a much smaller subset of the
same large group, i.e. a sample. Ideally, this is done in a way that allows valid conclusions to be
drawn about the population from the data obtained from the sample. Thus, a sample needs to be
adequately representative of the characteristics of interest of the population from which the sample
has been drawn.

A sample is nothing but a smaller representative of a larger population (Goode & Hatt, 1981). In
sampling terms, a population is any set of individuals who share a given set of characteristics, for
example, patients in a particular hospital, teachers of a particular school, or a group of people who
use the same supermarket. It is also called universe and may be finite or infinite. A finite population
is one where all the members can be counted easily, like the population of students in a particular
school; whereas fishes in a river cannot be counted easily, and thus are an example of an infinite
population, whose size is unlimited. While a measure based upon the entire population is called a
parameter, a measure based upon a sample is called a statistic.

According to Blalock (1960), a sample is a small subset of population selected via a probability
process. Further, he has categorised sampling methods into the following two categories:

1. Probability Sampling: These methods clearly specify the probability or the likelihood of
inclusion of each individual in the sample. They must satisfy three conditions, namely, the
size of the population must be known to the researcher; each individual must have an equal
chance of being included in the sample; and the desired sample size must be clearly
specified. The major advantage of the probability sampling methods is that these are free of
bias, in the sense that the obtained sample is considered representative. Hence, the
conclusions drawn from such samples are worth generalising to the entire population from
which they are drawn. The major disadvantage is that a certain amount of sampling error
exists because the researcher has only a limited no. of individuals in the entire population to
choose from. The smaller the sample, the greater the sampling error.
Following are the two major methods:

a. Simple random sampling: A simple random sample may be defined as one in which
each individual of the population has an equal chance of being included in the
sample, and also the selection of one individual is in no way dependent upon the
selection of another individual. For example, to draw a sample of 10 students from a
population of 40 students, their names can be written on slips of paper, placed in a
box and reshuffled thoroughly. The reshuffling gives all 40 students an equal chance
of being selected. This also ensures that the sample drawn is representative of the
population. As the sample size increases, it becomes more representative. However,
a large sample size is required to establish the reliability.
 b. Stratified random sampling: In this, population of interest is first divided into two or
more strata, which may be based upon a single criterion such as sex or a
combination of two or more criteria such as sex and graduation. These
subpopulations are non-overlapping, and internally homogeneous. Then, a simple
random sample for the desired number is taken from each subpopulation. This
process of stratification tends to increase the precision in estimating the attributes of
the whole population and gives some convenience in sampling by allowing
delegation of sampling. However, improper stratification can lead to wrong results.

c. Cluster sampling: This enables a representative sample to be generated when a

complete sampling frame is unavailable, but the population to be sampled can be
organised hierarchically, with each high level including all the lower levels. It is also
useful when the population is very large or dispersed geographically.

2. Non-probability sampling: With these methods, there is no way of assessing the probability
of each individual of the population being included in the sample. They provide no basis for
estimating how closely the characteristics of sample approximate the parameters of
population from which the sample has been drawn. Following are some important
techniques:

a. Accidental Sampling: Also known as convenience sampling, it is a procedure in

which the researcher selects cases according to their convenience. Such samples are
cheap, quick and highly unrepresentative. Thus, they can give a distorted view and
no generalisations can be made.

b. Quota Sampling: Similar to stratified random sampling, it identifies different strata of

population, for example, male and female; age related categories, and from each
stratum, a fixed number of individuals are selected arbitrarily. Since this technique
lacks randomness, the drawn conclusions lack external validity and generalisability.

c. Purposive Sampling: This technique is used in exploratory research or in field

research, wherein cases are selected with a specific purpose in mind. For example, if
the aim is to study the attitudes of British nationals towards refugees, it is important
that the sample is drawn from a specified population, i.e. British nationals. This
technique is convenient, however there is no way to judge the representativeness of
the sample.

d. Snowball Sampling: It is based on a sociometric, wherein all individuals of a group

identify their friends, who in turn identify their friends and associates, until a
network of friendships converges into some type of a definite social pattern. This
technique reveals communication pattern and helps in studying informal social
group. however, since the sample is not randomly drawn, some bias is introduced.

e. Systematic Sampling: It may be defined as selecting every nth individual from a

predetermined list of individuals. For example, selecting every 5 th roll number in a
class of 60 students. The first individual is chosen at random, however choosing
every nth individual makes this a non-probability technique. As relatively quick as
 this method may be, sampling error can happen if the list is arranged in a particular
order.

Meaning and relevance of qualitative research. Explain ethical guidelines (Block 4, Unit 1) done

The qualitative approach to research places importance on processes and meanings that cannot be
quantified. It provides a deeper understanding of the phenomenon within context (Guba & Lincoln,
1994). A qualitative research adopts a broadly constructivist/interpretivist approach to the social
world, although it is not based a single unified method of approach (Potter, 1996). It includes a broad
set of research methods and ways of analysing data. This type of research embraces naturalistic,
ethnographic, participant observational, case studies, diary method, discourse analysis, grounded
theory and several other methods. All these methods involve the collection and/or interpretation of
linguistic materials that can be subjected to analysis and interpreted for information about how
social experience is constructed in particular situations.

The goal is to comprehend the meanings attached to experiences by individuals, i.e. individuals’
interpretation of their own experiences in different situations. Thus, interpretation plays a key role in
qualitative research. It helps to bridge the gap between the researcher, the object being researched
and the way it has been represented. For example, if a researcher is studying a text based on a
schizophrenic episode, it would be important to remember that the text itself is an interpretation of
the actual episode by the writer. The actual events may or may not necessarily be exactly the same
as the text. In this sense, interpretation can only bridge these gaps, but not remove them.

Qualitative research is the most useful when there is limited theoretical information available about
the phenomenon that is under study as it may be impossible to develop precise hypotheses,
research questions or operational definitions. A qualitative approach in such cases is appropriate
because they can be exploratory (Sullivan, 2001).

 It seeks to confirm hypotheses using structured methods like in-depth interviews, participant
observations, experiences, etc.

 The questions used for data collection are usually open ended and can be hypothetical,
interpretative, multiple or leading so as to focus on the experiences and opinions of the
respondents.

 Qualitative research can be based on secondary data such as structured texts like stories,
news articles, books, etc. or primary data including unstructured texts like transcriptions for
interviews, focus group discussions; audio recordings, music and video recordings like
graphics, art, pictures, etc.

 The sample is usually selected via the method of purposive sampling that would allow
understanding the social phenomenon of interest.

 Furthermore, the research design in a qualitative research is relatively flexible, taking into
account the situational aspects.

 The next step is to analyse data that has been collected into some sort of explanation,
understanding or interpretation of the people and situations that are under study. Data
 analysis is based on an interpretative philosophy wherein the idea is to examine the
meaningful and symbolic content of qualitative data.

The data analysis can be pursued in different ways, to different degrees of depth and detail on the
basis of the purpose of research. At the simplest level, qualitative analysis involves asking questions
about the meaning of data and about the connections between its structure and content. If the
analysis is to be successful, it is critical to organise the data well due to its large volume and find a
time-efficient way to do it. Grounded theory, thematic analysis and discourse analysis are some of
the approaches to analyse qualitative data.

Thus, qualitative research is less interested in explaining phenomena than in understanding them. It
helps in attaining an in-depth knowledge of human nature, attitudes, behaviour and experiences. It
has the following implications in the field of psychology:

 Helps in textual description of people’s experiences

 Helps to identify and explain social norms, religions, gender roles and socio-economic status
 Helps to understand behavioural phenomena that cannot be quantified, like the
psychological impact of unemployment
 Helps in data collection in a more natural setting
 Helps in determining those factors that are meaningful and important to the respondents
from the sample. This is done by using open ended questions that provide a chance to
explore those facts that cannot be done with the help of close ended questions.

Following are some ethical guidelines that must be adhered to in qualitative research:

 The researcher must respect the respondents and their responses

 The researcher must also be respectful towards the community they are studying
 The respondents must be given adequate information about the research
 Confidentiality of respondents must be maintained at all times
 The researcher must be aware of the expected risks, if any, and benefits including
psychological and social, while conducting their research

Meaning and functions of research design. Types of single factor RDs (Block 3, Unit 1)

A research design is the detailed plan or a blueprint of the study that contains information about the
detailed procedure of testing the hypotheses and analysing the obtained data. It also operationalises
the variables so that they can be measured and helps in choosing the appropriate sample (Thyer,
1993). Thus, it may be defined as the sequence of steps to ensure that relevant data is collected in a
way that allows objective analysis of the different hypotheses formulated with respect to the
research problems. The selection of research design is based on the purpose of research, types of
variables and the conditions in which the study is to be conducted.

Based on the definition, following are the two functions of a research design:

 It provides the answers to research question as objectively, validly and economically as it is

possible. A research design guides the researcher on how to collect data for the formulated
hypotheses, which variables should be treated as control variables, what methods of
manipulation should be used in a particular context, what types of statistical analyses should
 be done and finally a possible answer to the research problem. Thus, it enables the
researcher to draw a valid and objective answer to the research problem.

 According to Kerlinger (2007), the main function of a research design is to control variance,
based on the MAXMINCON principle, i.e. maximise systematic variance, minimise error
variance and control extraneous variance.

o Systematic variance is the variance produced in the dependent variable by the

manipulation of the independent variables. The aim is to maximise this variance in
order to get as valid and objective data as possible.

o Error variance is defined as those variances or variabilities in the measures which

occur as a function of the factors not controlled by the experimenter/researcher.
Such factors can either be related to individual differences among the participants
themselves such as their attitude, motivation, etc. or they may be related to errors of
measurements such as the differences in the conditions of experiment, fatiguability,
etc.

o Extraneous variance is produced by the extraneous variables or the relevant

variables, i.e. those variables that are not being studied but could potentially affect
the dependent variable. The aim is to control this variance, and if possible,
completely eliminate it. For example, in a study being conducted to see if online
learning increases student understanding of statistics, it would be important to check
for the previous knowledge of the subject among the sample before starting the
study.

Research designs can be of different types, based on the purpose of research, the number and types
of variables. Single factor designs are used when there is only one independent variable. They can be
classified into the following two categories:

1. Between group design: In this, participants are randomly assigned to different treatment
conditions, to investigate the effect of different conditions on the participants. It is assumed
that assigning participants randomly to two or more groups will make these groups
statistically equivalent on the participant relevant variables (e.g., attitude, ability, etc.) which
may produce variations in the dependent variable. These can be of two types:

a. Two randomised group design:

i. In this, participants are randomly assigned to only two groups. For this, the
researcher first defines the independent variable and the dependent
variable, and the research population. Then, two values of the independent
variable are chosen, called ‘treatments’ or ‘conditions’. The aim is to examine
whether these two conditions affect the dependent variable in a differential
way.

ii. For example, a researcher wishes to study the effect of reward upon the rate
of learning verbal concepts among kindergarten children in a city. The
researcher randomly selects a sample of 100 children and will then randomly
 assign them to two groups, i.e. rewarded condition (experimental group) or
non-rewarded condition (control group). This random assignment can be
done by various methods.

iii. The most common method is the use of the Table of Random Numbers. For
this, the researcher may first list down all participants alphabetically and
then assign the 1st student to the experimental group, 2 nd student to the
control group, 3rd to the experimental group, and so on. The researcher
could also write the names on separate slips and pick them out, by assigning
every 2nd slip to the control group or vice versa. It is expected that these two
groups will not differ significantly at the start of the experiment.

iv. In the example, the children in experimental group will be rewarded for
learning the verbal concepts whereas the control group will not be rewarded
for the same task. Subsequently, the scores of all participants will be
recorded and subjected to statistical analysis, i.e. in this case, t test or Mann-
Whitney U test. If the two groups differ significantly on the DV, it can be
concluded that reward accelerates the learning of verbal concepts in
kindergarten children.

b. More than two randomised group design or multi group design: Due to some
limitations of two randomised group designs, more emphasis is given to the multi-
group design, wherein there are 3 or more conditions of the independent variable,
and accordingly, three or more groups of participants. For instance, in the
aforementioned example, the researcher could choose to have more than one
experimental group. The G1 would receive reward after every successful trial of the
task, while the G2 would receive reward after every 5 successful trials, and the G3
would be the control group that would not receive any reward. Thus, the
experimental group 1 and 2 differ in terms of the frequency of rewards. The scores
will be recorded and subjected to statistical analysis. In a multi-group design, the
most commonly applied statistic is ANOVA and Duncan Range test.

c. Matched group design: In this, first all subjects are tested on a common task
(matching variable) and then they are formed into groups (as many as needed) based
on their performance on the matching variable. The formed groups are known as
equivalent groups. Subsequently, the different conditions of the IV are introduced to
each group. If the groups have equivalent means on the DV before the experimental
treatment, and if a significant difference occurs after the treatment, the resulting
differences in the DV may be attributed to the experimental treatment. In this
design, two methods of matching are used:

i. Matching by pairs: Based on scores obtained by participants on the

matching variable, they are matched so that each person in group 1 has a
match in group 2.

ii. Matching in terms of mean and SD: When the basis of matching is the mean
and SD of the scores obtained on the matching variable, it is possible to form
 as many groups as needed in a way that these means and SDs do not
statistically differ.

2. Within-subject design: This is also known as repeated measure design because the same
individual is treated differently at different times and their scores are compared after they
have been subjected to different treatment conditions. It can be of two types:

a. Two conditions within-subject design: The same group of participants is tested

under two different conditions. The mean difference under these two conditions
may be tested through the match t test or Sandler’s A test. For example, to
investigate the effect of a drug upon retention of a verbal task, participants can be
subjected to two conditions, i.e. normal condition when the group has not taken any
drug and a condition where it is under the influence of a drug.

b. Multiple conditions within-subject design: This involves subjecting the same group
of participants to three or more conditions. For example, in the field of proactive
inhibition, in condition I the group may be required to learn zero list; in condition II
the group may be required to three lists; in condition III the group may be required
to learn four lists; in condition IV five lists; and in condition V seven lists. At the end
of each condition the group may be required to learn one new list. After an interval
of one-hour post learning one new list, the group may be asked to recall items from
the new list. It is expected that the no. of prior lists learnt will proactively inhibit the
retention of new list.

c. Controlling for order and sequence effects:

i. Order effect is the influence of a particular trial that arises from its position
in the sequence of trials and sequence effect arises from the interaction
between specific trials of the experiment. For example, when judging the
heaviness of lifted weights, there is a possibility that light weight will feel
lighter if it is followed by a heavy one, and vice versa.

ii. Controlling for these effects can be done by randomisation of trials. For
example, if there are four conditions that are to be presented twice, they can
follow the sequence BCAD, ADCB.

iii. When there are fewer subjects, then several conditions can be presented a
few times only. In such cases, counterbalancing can be used to control order
effect. For example, three conditions ABC can be presented as ABC, CBA
sequence.

Comparison between between-group and within-group design

 Since the measures are repeated on the same group of subjects in a within-group design, the
factor of individual differences is automatically controlled. This increases correlation among
measures, reducing the error variance. In a between-group design, the problem of individual
differences remain since there are two groups of subjects. This can increase the error
variance.
  Within-group designs are especially needed for those areas which require studying the
changes in participants’ behaviour over time. Thus, panel and longitudinal studies are most
appropriate for such designs.

 When there are chances of practice or carry over effect of one treatment to the subsequent
task, then between-group design should be preferred.

Section B (400 words)

Advantages and disadvantages of quasi experimental design (Block 3, Unit 3)

Quasi-experimental design is the one that resembles a true experimental research design, but is not
exactly the same. Although, sometimes, the independent variable is manipulated, participants are
not randomly assigned to conditions or groups (Cook & Campbell, 1979). Since there is no or very
little randomisation, it is likely that the outcome is being affected by confounding variables. This
poses a major threat to its internal validity. However, as long as the shortcomings of this design are
recognised, it can be a useful tool in situations where true experiments are not possible.

Some of the advantages and disadvantages of a quasi-experimental design are given below:

Advantages:

 Quasi experimental design can be useful when pre-selection and randomisation of groups is
not possible. For example, to study the effects of alcohol consumption during pregnancy on
child’s development, it would not be possible to randomly assign mothers to group. Those
mothers would have to be chosen who consumed alcohol during pregnancy vs. those who
didn’t.
 Such designs may reduce the time and resources required because the process of pre-
selection and randomisation is not required.
 These designs are relatively easier to set up than true experimental designs.
 External validity is increased in quasi-experimental designs as such research takes place in a
natural setting and thus the responses are likely to be more genuine.
 These designs can also be very useful in identifying general trends from the results.
 A quasi-experimental design can be efficient for longitudinal research that involves longer
time periods, followed up in different environments.

Disadvantages:

 The lack of random assignment into test groups can lead to non-equivalent groups, which
may limit the generalizability of the results to a larger population.
 Statistical analysis may not be meaningful due to the lack of randomisation and threats to
internal validity.
 The lack of randomisation may make the study more feasible, but it makes it harder to
control the effect of confounding variables. For example, when examining the impact of
smoking by pregnant mothers, there may be other factors such as diet, education, overall
health and access to health care in general that may influence the results.
 Establishing causality is difficult with such designs due to a variety of extraneous and
confounding variables that exist in a natural, social environment.
Explain goals, methods and steps of grounded theory (Block 4, Unit 2)

Grounded theory is an approach to textual materials that involves systematically reading the text to
identify the categories and concepts that are expressed in it. It provides a direction to the researcher
and directs them to generate new theories or modify existing ones.

Following are the basic goals of grounded theory:

 Grounded theory assures to provide a good theory as it is based on data collected from
multiple sources.
 This approach places a great deal of importance on the quality of theory through constant
evaluation of theories.
 It allows formulation of hypotheses based on conceptual ideas.
 It lets the research to discover the participants’ main concern and how they can continually
try to resolve it.
 It aims to generate those concepts that can explain behaviour regardless of time and place.

The method of ground theorists involves anything that can be treated as relevant ‘data’, for example,
field noting interviews, lectures, seminars, etc. Anything that can help the researcher generate
concepts for the emerging theory is considered data. Sometimes the researcher may interview
themselves as it can provide insight into researcher’s own knowledge at the conceptual level.

Following steps are involved in grounded theory approach:

 Memoing: the first step is to collect data in form of memos, i.e. short notes that researcher
writes and prepares. These memos can act as a source of data which is further put in other
processes of analysis and interpretation. Memos can be prepared in the following way:
o Theoretical note: It includes details on how a particular source of data/text is related
to the existing literature of the concerned study and is usually one to five pages long.
o Field note: It includes notes prepared based on observations or interactions while
actively participating with the population or community that is under study.
o Code notes: These notes discuss the codes of categorised things, properties and
events. They can also guide the theorists while analysing a text or case.
 Sorting: Once memoing is completed, the collected data is organised to form a proper
linkage of information and ideas. This step can also help reveal new information that was
not revealed in the previous stage.
 Writing: Once the collected information is organised, the theorists begin writing the report
that gives a meaningful shape to the data. This is similar to preparation of research reports,
that link the collected information with the already existing literature and provide future
directions for the theory.

Discuss purpose, types and steps of ethnographic research method (Block 4, Unit 1)

Ethnographic research method intends to study culture through close observation and active
participation. It is generally used as a way of developing a deep understanding of the everyday life of
a social group from the perspective of its members. By actively participating in the group and sharing
the experiences of others allows the researcher to approach the understanding of a cultural insider
to a degree that would otherwise be impossible to achieve. The data is collected using chain
sampling in all empirical areas of investigation. Such a study requires the researcher to stay within
the community over a considerable period of time to obtain full acceptance by the other
participants. The analysis of data leads to development of theories for the socio-cultural phenomena
that’s under study.

Following is the purpose of an ethnographic study:

 Helps in cross-cultural analysis

 Helps in analysing the past events or the history of the culture
 Helps in studying individual behaviour in their natural environment
 Such close observation of the group increases the validity of reports and formulated theory

Following are some of the types of ethnographic research:

 Macro ethnography: it is the study of a broadly defined cultural group like the Turkish.
 Micro ethnography: it is the study of a more specific cultural groupings like the local
government.
 Emic perspectives: under this approach, the responses of the ingroup members are noted
down.
 Etic perspectives: the responses of the outgroup members are noted down.

Following are the steps involved in an ethnographic study

 Selection: the study begins with the selection of a social group of interest that the researcher
wants to study.
 Review of literature: this involves reviewing the existing literature about the social group to
get a brief idea of their history.
 Identification of variables: the variables are chosen based on both the researcher’s and the
group members’ interest.
 Entry: the researcher tries to penetrate the culture to gain acceptance from the group
members.
 Cultural immersion: during this period, the researcher lives within the community for a
period of time, trying to gain the confidence of the group members.
 Data collection: the data is collected in form of observational transcripts and interview
recordings and tapings.
 Development of theory: the researcher formulates theory on the basis of interpretation of
the results.

Define correlational RD and discuss its type (Block 3, Unit 4)

It is a quantitative design that attempts to compare the level of one variable with those of another,
to see if a relationship exists between the two. In the correlational research design, the researcher
collects two or more sets of data from the same group of subjects to determine the relationship
between the two data sets (Singh 1998). It is a non-experimental design, i.e. it does not involve the
manipulation of the IV to see its effect on DV. Thus, a significant correlation between two variables
may not necessary mean that one variable causes the other one.

For example, Segall et al. (1963) compared the perception of geometric illusions in different national
and ethnic groups that included Black African, Filipinos, white South Africans, and white Americans.
The findings suggested that members of black African and Filipino groups were much less susceptible
to the illusions than the others. In this case, it is difficult to suggest a causal relationship between
ethnicity and susceptibility to illusions.

Following are the two types of correlational research design:

1. To measure the degree of association between variables:

a. Association between two variables
For example, a researcher wants to study the relationship between academic
procrastination and academic achievement of college students. For this, data is
collected from 100 college students who are randomly selected, using the measures
of academic procrastination and academic achievement. This gives the researcher
two sets of data whose correlation can be calculated. The correlation value could
range from -1.00 to +1.00. The larger the correlation coefficient, the more consistent
the association between the variables.
A positive correlation coefficient implies a direct relationship between the two
variables. Thus, as variable 1 goes up, so does variable 2. A negative correlation
coefficient indicates an inverse relationship, i.e. both variables move in the opposite
direction.
b. Association between more than two variables
In the same example, if other variables are to be included, like intelligence, hope,
etc., the researcher could also measure the association between all these variables
in the same sample. In this case, multiple correlation coefficients will be obtained.
2. To predict the score on one variable using knowledge about one or more variables:
If there is a correlation between two variables, the score on variable 2 can be predicted if the
score on variable 1 is known. It is possible to have multiple predictor variables and for one
criterion variable. The prediction of scores is done through regression analysis.

Discuss the steps and types of survey research (Block 2, Unit 1)

Following steps are involved in a survey research:

1. Determination of the aims and objectives of the study: To make the purpose of research
relevant and understandable, it is important to clearly define the basic aims and objectives of
the study.
2. Define the population to be studied: The next step is to define the target population that will
be studied.
3. Design and construct a survey: A pilot study can be conducted to analyse the significance and
relevance of the survey that is constructed along with the relevant questions and other
instruments.
4. Select a representative sample: A representative sample from the target population is chosen
so that the findings can be easily generalised on the population.
5. Administer the survey: The data collection from the sample begins.
6. Analyse and interpret the findings of the survey: The collected data is analysed using the
appropriate statistical tools and the findings are then interpreted based on the results.
7. Prepare the report of the survey: A report is prepared, containing all the details of the
previous steps, to evaluate the findings in relation to the proposed aims and objectives of
the research.
 8. Communicate the findings: It is important to disseminate the findings to the target
population, to maintain a record for the future research and/or to inform policy decision-
making.

Following are the two broad types of survey research:

1. Cross-sectional survey:
A cross-sectional survey is time and money efficient, as data is collected from different types
of groups at a single time. For example, a cross-cultural study.
2. Longitudinal survey:
This survey is used when the same sample is to be studied for behaviour, attitudinal or
developmental changes, over a period of time. They can be of the following types:
a. Trend Studies: These are used to analyse a trend of a phenomenon in a
population. In this case, the sample may not be the same over time, but the
population will always be the same. For example, a yearly survey of graduate
students to study the employment rate of the university.
b. Cohort Studies: Such studies involve a particular sample (cohort) that is studied
repeatedly over a period of time. For example, tracing emotional development
of a cohort at different developmental stages.
c. Panel Studies: These studies also involve the same sample (panel) every time
and are used to investigate changes in behaviour, attitudes or practices within a
period. For example, tracing changes in academic procrastination in school
students over the years.

Section C (50 words)

Strategies of data interpretation in qualitative research (Block 4, Unit 4)

Following are the ways in which data can be interpreted in qualitative research:

 Making a list: Summarising the categories and coding them to make a final list of findings
(which can be used later for explanation while writing the report.)
 Elaborative narratives: Elaborating on the information collected via interviews, recordings.
 Use of matrices: A tabular or a diagrammatic representation of data (flowchart) can be used
to compare different groups or data sets on important variables.

Implications of content analysis (Block 4, Unit 3)

Content analysis can have the following implications:

 Helps in establishing a link between cause and effect

 Evaluate and improve the programming of the media world
 Increase awareness and summarise the notes that focus on specific issues
 Aids in making inferences of the causes

Field experiment (Block 2, Unit 3)

Field experiments are conducted in any non-lab environment, wherein people engage in the normal
tasks of everyday life- such as working, playing, etc. and thus, behaviour can be observed under
natural conditions. Different situations are created in the field to represent different levels of IV and
the participants’ unconstrained responses are observed. (Reduces the likelihood of reactivity effects
because the participants are unaware that the experiment is taking place. Arranging a satisfactory
field experiment may be difficult as the researcher has no control over the events that take place and
there may be ethical issues around the need to obtain informed consent.)

Reliability (Block 1, Unit 2)

Reliability is one of the most important characteristics of any test and it refers to the precision, or
accuracy, of the measurement or score. It is the consistency of scores or measurement which is
reflected in the reproducibility of the scores. The higher the reliability of a test, relatively the freer it
would be of measurement errors.

Types of case study (Block 2, Unit 4)

 Illustrative case study: Primarily descriptive, these studies typically utilise one or two
instances of an event to describe a situation. (These aim to make the unfamiliar familiar.)
 Exploratory case study: These studies are conducted before implementing a large-scale
investigation, to identify questions and select types of measurement.
 Cumulative case study: It aggregates information from past studies to allow greater
generalisation without additional cost or time.
 Critical Instance case study: It is used for answering cause and effect questions.

Types of hypotheses (Block 1, Unit 4)

Hypotheses can be broadly categorised into the following:

Null hypotheses (H0) (is used to test the significance of difference.) It asserts that there is no true
difference between two population means and any difference found is purely accidental and
unimportant.

Alternate hypotheses (Ha) is the opposite of Ho, i.e. it asserts that a difference is expected between
the two population means. (This assertion can either be directional, i.e. specifying which population
mean is expected to be higher/lower on the variable interest or non-directional, i.e. not making any
predictions about the direction of difference.)

Constructs (Block 1, Unit 3)

Constructs are new concepts for scientific purposes, i.e. it is invented specifically for a special
scientific purpose. They generally have two common characteristics: they are a part of a theoretical
framework and is related in various ways to other constructs; and, they are usually operationally
defined so as to allow its observation and measurement. (For example, reinforcement- theoretically,
it is related to other constructs such as drive, motivation, etc.; operationally, it may be defined as any
stimulus which increases the probability of the occurrence of a (desired) response.)

Qualities of a good research (Block 1, Unit 1)

 Systematic: Research is structured according to a set of rules in a specified sequence.

 Empirical: Conclusions drawn are based upon evidence gathered from data collected from
real life experiences and observations.
  Valid and verifiable: Reliable and valid instruments are used for data collection and results
are obtained using statistical measures.
 Logical: Follows either inductive (general to specific) or deductive (specific to general)
approach.
 Develops theories and principles: Concludes with sound generalisations regarding the whole
population.
 Replicable: The research design, procedures and results should be replicable to assess its
validity.

Advantages of a factorial research design (Block 3, Unit 2)

 Enables the manipulation and control of two or more IVs simultaneously.

 More precise than single factor design (Kerlinger, 2007).
 Helps to separate the main effects of IVs and interactive effect of two or more IVs.
 Results are more comprehensive and can be generalised due to the manipulation of several
IVs in a single experiment.

Research biases (Block 1, Unit 1)

Following biases can distort the perceptions about collected data:

 External influences like culture or the media.

 Personal biases like personal beliefs, attitudes or past experiences.
 Observer biases must be kept aside while data collection and the aim should be to remain as
objective as possible.
 Expectancy bias (affects observations by), i.e. encouraging reactions to the events being
observed.
 Placebo biases (operate when people strongly want to believe), i.e. a strong belief that a
treatment is successful.