Cross-Sectional, Longitudinal, and Case Study Designs Chapter 7
Cross-Sectional, Longitudinal, and Case Study Designs
Chapter 7
Introduction
Study Design A research strategy specifying the number of
cases to be studied, the number of times data will be collected, the number of samples that will be used, and whether or not the researcher will try to control or manipulate the independent variable in some way
Introduction
Connections Between the Uses of Research, Theory, & Study Design
Decisions about study design are based partly on
Uses or purposes of the research The researcher's interest in testing a causal
hypothesis Use of theory
Study Design Choices
Cross-Sectional Study Design Cross-Sectional Studies and Causal
Relationships Longitudinal Designs: Panel, Trend, &
Cohort Designs
Study Design Choices
Cross-Sectional Study Design A study design in which data are collected for
all the variables of interest using one sample at one time
Most frequently used design in social science research
Study Design Choices
Cross-Sectional Studies and Causal Relationships Cross-sectional designs are sometimes used
to examine causal hypotheses. Remember from Chapter 2 – just because two
variables “go together” doesn’t mean that one variable causes change in another variable
Study Design Choices
Cross-Sectional Studies and Causal Relationships Causal Relationship
A genuine relationship between an independent and dependent variable with the independent variable occurring before the dependent variable
Study Design Choices
Cross-Sectional Studies and Causal Relationships
Requirements for a causal relationship1. Empirical association between the independent
and dependent variables2. Temporal precedence or time order – the
independent variable occurs first, thus causing the dependent variable
3. False relationship – be certain that the action is not caused by a third variable – antecedent variable
Study Design Choices
Longitudinal Designs A research design in which data are collected
at least two different times, such as a panel, trends, or cohort study
Used to track changes over time, rather than relying on a cross-sectional design asking retrospective questions
Allows for more analysis, by treating each data set as if it were cross-sectional and it is possible to do over-time analyses
Study Design Choices
Longitudinal Designs Panel Study
A study design in which data are collected about one sample at least two times where the independent variable is not controlled by the researcher.
Requires following one sample over time
Study Design Choices
Longitudinal Designs Panel study
Panel attrition The loss of subjects from a study because of
disinterest, death, illness, or inability to locate them
Study Design Choices
Longitudinal Designs Panel study
Panel Conditioning The effect of repeatedly measuring variables on
members of a panel study Participants tend to become more conscious of their
attitudes, emotions, and behavior with repeated data collections
Study Design Choices
Longitudinal Designs Trend study
A study design in which data are collected at least two times with a new sample selected from a population each time.
Useful when researchers are interested in identifying changes over time in a large population
Avoids the time and expense of relocating the same individuals to track changes
Study Design Choices
Longitudinal Designs Cohort study
A longitudinal study which follows a cohort over time
Study Design Choices
Longitudinal Designs Cohort study
Cohort A group of people born within a given time frame or
experiencing a life event, such as marriage or graduation from high school, in the same time period.
Study Design Choices
• Longitudinal Designs Cohort study
Birth cohorts are the most frequently selected cohort People can exit a cohort (by dying) but no one can
enter after it is created Either panel or trend study design may be used with
a cohort study
Focal Research
“Moving On? Continuity and Change after Retirement” Adler and Clark collected data twice from the
same sample and therefore went beyond the cross-sectional design.
Why do you think they didn’t use a cross-sectional study design?
Focal Research
Thinking about ethics Submitted our research to our university’s Institutional
Review Board (IRB) for approval. The participants volunteered to be part of the study by
responding to the request for participation, they also voluntarily agreed to the second interview and each time gave informed consent before being interviewed.
Believe that no harm came to any of the participants as a result of being part of the study, we have kept all names and information confidential and, in all reports, we have identified none of the participants by their real names
The Case Study
Case study A research strategy which focuses on one
case (an individual, a group, an organization, etc.) within its social context at one point in time, even if that one time spans months or years.
A design with a long and respected history Generalizability is limited, due to relatively
small numbers of participants
The Case Study
Case study Holistic approach is the distinguishing feature
of a case study Looking at the case as a whole embedded in its
natural context and using an open-ended rather than prestructured strategy for collecting data.
Relies on several data sources, is conducted in great detail, and results in an in-depth, multifaceted investigation of a single social phenomenon
Summary
The purposes or reasons for the research influence the choice of study design
Questions to think about Is the focus on variables and the connections between them, or
is the goal to analyze only one case or a small number of cases holistically?
Is it useful and possible to collect data more than one time from the same sample?
Is there a cohort that can be studied to answer the research question?
Is it useful and possible to select a new sample each time if data will be collected more than once?
If there is a causal hypothesis, is it possible, useful, and ethical to try to control or manipulate the independent variable in some way?
Quiz – Question 1
A study examining how attitude about smoking influence other health behaviors that will survey a random sample of city residents during the month of August, is an example of
a. Longitudinal research
b. Experimental research
c. Cross-sectional research
d. Case study research
e. Questionnaire research
Quiz – Question 2
If we are interested in establishing that two variables are related in a causal way (A causes B), three conditions must be satisfied. Among the following items, which is not a factor in this determination?
a. A and B must be the result of another variable, C.
b. A must occur prior to B
c. As A changes, so does B
d. If B changes, A will change as well
Quiz – Question 3
If we are interested in how individuals’ attitudes about smoking marijuana change from adolescence to adulthood, marriage and parenthood, the most appropriate study design would be
a. Longitudinal
b. Case study
c. Cross-sectional
d. Surveys
e. Questionnaires
Quiz – Question 4
If we are interested in learning about how social and economic elites influence decision making in one city and will rely upon census data, newspaper accounts, interviews, and surveys to do this research, we are relying on the ________________ study design.
a. Caseb. Non-experimentalc. Trendd. Longitudinale. Questionnaire
Quiz – Question 5
If you were conducting research using a cross-section design with retrospective questions about childhood and adolescent years to young people in their twenties, you
a. should yield consistent resultsb. may find this problematic for the accuracy of relying on people’s
answers about what they felt or did years beforec. will have problems with internal validityd. b & ce. None of the above