Page 79 - Academic and Research Book_WEB

preferably done using random sampling. The National Child Development Study (NCDS) of England
is an example of a cohort study.

Case Study Design

This design is suited more to studies in sociology, psychology, and anthropology. It is appropriate in
situations where the attempt is to study in detail, a single community (including online communities),
a single school of thought, a single family, a single organization, a person (such studies often make use
of the life or oral history, or biographical approach), or a single event. These entities of study on which
a case study design can be applied are generally called “social settings”. A case study design is used for
a detailed, intensive study of the complexity in nature of one particular social setting. Cases could be
categorized into: a critical case, an extreme or unique case, a representative or typical case, a revelatory
case, or a longitudinal case.

Comparative Design

A comparative design is appropriate when two or more contrasting social settings or cases are
studied using identical methods, at almost the same point in time. This design is also suited to
cross-cultural and cross-national comparative studies. The issues of reliability, validity, replicability,
and generalizability of comparative designs are similar to those of the cross-sectional design. A
comparative design gives an opportunity to the researcher to examine the causal mechanisms
in contrasting or similar situations or contexts. It is a variant of the comparative design, known as
multiple-case study approach. The multiple-case study approach is suitable to qualitative case studies
as well.

Data Collection Methods

Social scientists have, of late, been making a distinction between: (1) naturally occurring data, and
(2) non-naturally occurring data. Conversational analysts and discursive psychologists strongly
recommend working with recordings of naturally occurring data and their transcripts (Speer, 2008).
The methods of collecting non-natural or contrived data are discussed as follows:

Self-administered Questionnaires and Standardized Interviews

Self-administered questionnaires are handed over to the respondents, for reading and filling in
their responses. They are characterized by the complete absence of the interviewer. This gives rise
to concerns about non-responses, item non-responses, and the quality of data. This is because the
question and its response alternatives would be interpreted differently by different respondents.
In contrast to that, in the case of structured interviews, the presence of the interviewer can help in
explaining the questions and the response alternatives. This enhances the data quality and reduces
item non-response. Moreover, the non-response rate in case of structured interviews would be
relatively less because the interviewer may successfully convince the reluctant respondents, who
would, in the absence of the interviewer, might be non-responsive. Standardized interviews are better
to get responses on sensitive questions. On the other hand, interviewer effects would have to be
accounted for, in the case of standardized interviewing (De Leeuw, 2008).
In case of self-administered questionnaires, the respondents see the questions and the response
alternatives themselves. Self-administered questionnaires can be designed using online services such

68 | University Grants Commission