Professional Documents
Culture Documents
edu/writingguide/researchdesigns
The research design refers to the overall strategy that you NOTE: To search for scholarly resources on specific
choose to integrate the different components of the research designs and methods, use the SAGE Research
study in a coherent and logical way, thereby, ensuring Methods database. The database contains links to more
you will effectively address the research problem; it than 175,000 pages of SAGE publisher's book, journal, and
constitutes the blueprint for the collection, measurement, reference content on quantitative, qualitative, and
and analysis of data. Note that your research problem mixed research methodologies. Also included is a
determines the type of design you should use, not the collection of case studies of social research projects that
other way around! can be used to help you better understand abstract or
complex methodological concepts.
De Vaus, D. A. Research Design in Social Research. London: SAGE, 2001; Trochim,
William M.K. Research Methods Knowledge Base. 2006.
De Vaus, D. A. Research Design in Social Research. London: SAGE, 2001; Gorard, Stephen. Research
Design: Creating Robust Approaches for the Social Sciences. Thousand Oaks, CA: Sage, 2013; Leedy,
Paul D. and Jeanne Ellis Ormrod. Practical Research: Planning and Design. Tenth edition. Boston, MA:
General Structure and Writing Style Pearson, 2013; Vogt, W. Paul, Dianna C. Gardner, and Lynne M. Haeffele. When to Use What Research
Design. New York: Guilford, 2012.
1. Identify the research problem clearly and justify 1. This is a collaborative and adaptive research
its selection, particularly in relation to any valid design that lends itself to use in work or
alternative designs that could have been used, community situations.
2. Review and synthesize previously published 2. Design focuses on pragmatic and solution-driven
literature associated with the research problem, research outcomes rather than testing theories.
3. Clearly and explicitly specify hypotheses [i.e., 3. When practitioners use action research, it has the
research questions] central to the problem, potential to increase the amount they learn
4. Effectively describe the data which will be consciously from their experience; the action
necessary for an adequate testing of the research cycle can be regarded as a learning
hypotheses and explain how such data will be cycle.
obtained, and 4. Action research studies often have direct and
5. Describe the methods of analysis to be applied obvious relevance to improving practice and
to the data in determining whether or not the advocating for change.
hypotheses are true or false. 5. There are no hidden controls or preemption of
direction by the researcher.
The organization and structure of the section of your
paper devoted to describing the research design will What these studies don't tell you?
vary depending on the type of design you are using.
However, you can get a sense of what to do by 1. It is harder to do than conducting conventional
reviewing the literature of studies that have utilized the research because the researcher takes on
same research design. This can provide an outline to
follow for your own paper.
RESEARCH DESIGNS Page 1 of 10
http://libguides.usc.edu/writingguide/researchdesigns
Case Studies. Writing@CSU. Colorado State University; Anastas, Jeane W. Research Design for Social
Work and the Human Services. Chapter 4, Flexible Methods: Case Study Design. 2nd ed. New York:
II. Case Study Design Columbia University Press, 1999; Gerring, John. “What Is a Case Study and What Is It Good
for?” American Political Science Review 98 (May 2004): 341-354; Greenhalgh, Trisha, editor. Case Study
Evaluation: Past, Present and Future Challenges. Bingley, UK: Emerald Group Publishing, 2015;
Definition and Purpose Mills, Albert J., Gabrielle Durepos, and Eiden Wiebe, editors. Encyclopedia of Case Study Research.
Thousand Oaks, CA: SAGE Publications, 2010; Stake, Robert E. The Art of Case Study Research. Thousand
Oaks, CA: SAGE, 1995; Yin, Robert K. Case Study Research: Design and Theory. Applied Social Research
A case study is an in-depth study of a particular research Methods Series, no. 5. 3rd ed. Thousand Oaks, CA: SAGE, 2003.
Nonspuriousness -- a relationship between two same or similar characteristics that are relevant to the
variables that is not due to variation in a third research problem being investigated, rather than
variable. studying statistical occurrence within the general
population. Using a qualitative framework, cohort studies
generally gather data using methods of observation.
Cohorts can be either "open" or "closed."
What do these studies tell you?
Open Cohort Studies [dynamic populations, such
1. Causality research designs assist researchers in as the population of Los Angeles] involve a
understanding why the world works the way it population that is defined just by the state of
does through the process of proving a causal link being a part of the study in question (and being
between variables and by the process of monitored for the outcome). Date of entry and
eliminating other possibilities. exit from the study is individually defined,
2. Replication is possible. therefore, the size of the study population is not
3. There is greater confidence the study has internal constant. In open cohort studies, researchers can
validity due to the systematic subject selection only calculate rate based data, such as,
and equity of groups being compared. incidence rates and variants thereof.
Closed Cohort Studies [static populations, such
What these studies don't tell you? as patients entered into a clinical trial] involve
participants who enter into the study at one
defining point in time and where it is presumed
1. Not all relationships are casual! The possibility
that no new participants can enter the cohort.
always exists that, by sheer coincidence, two
Given this, the number of study participants
unrelated events appear to be related [e.g.,
remains constant (or can only decrease).
Punxatawney Phil could accurately predict the
duration of Winter for five consecutive years but,
the fact remains, he's just a big, furry rodent]. What do these studies tell you?
2. Conclusions about causal relationships are
difficult to determine due to a variety of 1. The use of cohorts is often mandatory because a
extraneous and confounding variables that exist randomized control study may be unethical. For
in a social environment. This means causality can example, you cannot deliberately expose
only be inferred, never proven. people to asbestos, you can only study its effects
3. If two variables are correlated, the cause must on those who have already been exposed.
come before the effect. However, even though Research that measures risk factors often relies
two variables might be causally related, it can upon cohort designs.
sometimes be difficult to determine which 2. Because cohort studies measure potential causes
variable comes first and, therefore, to establish before the outcome has occurred, they can
which variable is the actual cause and which is demonstrate that these “causes” preceded the
the actual effect. outcome, thereby avoiding the debate as to
which is the cause and which is the effect.
3. Cohort analysis is highly flexible and can provide
insight into effects over time and related to a
Beach, Derek and Rasmus Brun Pedersen. Causal Case Study Methods: Foundations and Guidelines for variety of different types of changes [e.g., social,
Comparing, Matching, and Tracing. Ann Arbor, MI: University of Michigan Press, 2016; Bachman,
Ronet. The Practice of Research in Criminology and Criminal Justice. Chapter 5, Causation and cultural, political, economic, etc.].
Research Designs. 3rd ed. Thousand Oaks, CA: Pine Forge Press, 2007; Brewer, Ernest W. and Jennifer
Kubn. “Causal-Comparative Design.” In Encyclopedia of Research Design. Neil J. Salkind, editor.
4. Either original data or secondary data can be
(Thousand Oaks, CA: Sage, 2010), pp. 125-132; Causal Research Design: Experimentation. Anonymous
SlideShare Presentation; Gall, Meredith. Educational Research: An Introduction. Chapter 11,
used in this design.
Nonexperimental Research: Correlational Designs. 8th ed. Boston, MA: Pearson/Allyn and Bacon, 2007;
Trochim, William M.K. Research Methods Knowledge Base. 2006.
What these studies don't tell you?
variables change during the course of the study, relatively inexpensive and take up little time to
potentially impacting the validity of the findings. conduct.
3. Due to the lack of randominization in the cohort
design, its external validity is lower than that of What these studies don't tell you?
study designs where the researcher randomly
assigns participants.
1. Finding people, subjects, or phenomena to study
that are very similar except in one specific
variable can be difficult.
2. Results are static and time bound and, therefore,
Healy P, Devane D. “Methodological Considerations in Cohort Study Designs.” Nurse
Researcher 18 (2011): 32-36; Glenn, Norval D, editor. Cohort Analysis. 2nd edition.
give no indication of a sequence of events or
Thousand Oaks, CA: Sage, 2005; Levin, Kate Ann. Study Design IV: Cohort reveal historical or temporal contexts.
Studies.Evidence-Based Dentistry 7 (2003): 51–52; Payne, Geoff. “Cohort Study.” In The
SAGE Dictionary of Social Research Methods. Victor Jupp, editor. (Thousand Oaks, CA:
3. Studies cannot be utilized to establish cause and
Sage, 2006), pp. 31-33; Study Design 101. Himmelfarb Health Sciences Library. George effect relationships.
Washington University, November 2011; Cohort Study. Wikipedia.
4. This design only provides a snapshot of analysis so
there is always the possibility that a study could
have differing results if another time-frame had
V. Cross-Sectional Design been chosen.
5. There is no follow up to the findings.
Definition and Purpose
What these studies don't tell you? 1. The design is artificial, and results may not
generalize well to the real world.
1. The results from a descriptive research cannot be 2. The artificial settings of experiments may alter the
used to discover a definitive answer or to behaviors or responses of participants.
disprove a hypothesis. 3. Experimental designs can be costly if special
2. Because descriptive designs often utilize equipment or facilities are needed.
observational methods [as opposed to 4. Some research problems cannot be studied
quantitative methods], the results cannot be using an experiment because of ethical or
replicated. technical reasons.
3. The descriptive function of research is heavily 5. Difficult to apply ethnographic and other
dependent on instrumentation for measurement qualitative methods to experimentally designed
and observation. studies.
Anastas, Jeane W. Research Design for Social Work and the Human Services. Chapter 7, Flexible
Anastas, Jeane W. Research Design for Social Work and the Human Services. Chapter 5, Flexible
Methods: Experimental Research. 2nd ed. New York: Columbia University Press, 1999; Chapter 2:
Methods: Descriptive Research. 2nd ed. New York: Columbia University Press, 1999; Given, Lisa M.
Research Design, Experimental Designs. School of Psychology, University of New England, 2000; Chow,
"Descriptive Research." In Encyclopedia of Measurement and Statistics. Neil J. Salkind and Kristin
Siu L. "Experimental Design." In Encyclopedia of Research Design. Neil J. Salkind, editor. (Thousand Oaks,
Rasmussen, editors. (Thousand Oaks, CA: Sage, 2007), pp. 251-254; McNabb, Connie. Descriptive
CA: Sage, 2010), pp. 448-453; "Experimental Design." In Social Research Methods. Nicholas Walliman,
Research Methodologies. Powerpoint Presentation; Shuttleworth, Martyn. Descriptive Research Design,
editor. (London, England: Sage, 2006), pp, 101-110; Experimental Research. Research Methods by
September 26, 2008. Explorable.com website.
Dummies. Department of Psychology. California State University, Fresno, 2006; Kirk, Roger E. Experimental
Design: Procedures for the Behavioral Sciences. 4th edition. Thousand Oaks, CA: Sage, 2013; Trochim,
William M.K. Experimental Design. Research Methods Knowledge Base. 2006; Rasool,
Shafqat. Experimental Research. Slideshare presentation.
2. Exploratory research is flexible and can address 3. Historical records can add important contextual
research questions of all types (what, why, how). background required to more fully understand
3. Provides an opportunity to define new terms and and interpret a research problem.
clarify existing concepts. 4. There is often no possibility of researcher-subject
4. Exploratory research is often used to generate interaction that could affect the findings.
formal hypotheses and develop more precise 5. Historical sources can be used over and over to
research problems. study different research problems or to replicate
5. In the policy arena or applied to practice, a previous study.
exploratory studies help establish research
priorities and where resources should be What these studies don't tell you?
allocated.
1. The ability to fulfill the aims of your research are
What these studies don't tell you? directly related to the amount and quality of
documentation available to understand the
1. Exploratory research generally utilizes small research problem.
sample sizes and, thus, findings are typically not 2. Since historical research relies on data from the
generalizable to the population at large. past, there is no way to manipulate it to control
2. The exploratory nature of the research inhibits an for contemporary contexts.
ability to make definitive conclusions about the 3. Interpreting historical sources can be very time
findings. They provide insight but not definitive consuming.
conclusions. 4. The sources of historical materials must be
3. The research process underpinning exploratory archived consistently to ensure access. This may
studies is flexible but often unstructured, leading especially challenging for digital or online-only
to only tentative results that have limited value to sources.
decision-makers. 5. Original authors bring their own perspectives and
4. Design lacks rigorous standards applied to biases to the interpretation of past events and
methods of data gathering and analysis because these biases are more difficult to ascertain in
one of the areas for exploration could be to historical resources.
determine what method or methodologies could 6. Due to the lack of control over external variables,
best fit the research problem. historical research is very weak with regard to the
demands of internal validity.
7. It is rare that the entirety of historical
documentation needed to fully address a
Cuthill, Michael. “Exploratory Research: Citizen Participation, Local Government, and Sustainable research problem is available for interpretation,
Development in Australia.” Sustainable Development 10 (2002): 79-89; Streb, Christoph K. "Exploratory
Case Study." In Encyclopedia of Case Study Research. Albert J. Mills, Gabrielle Durepos and Eiden therefore, gaps need to be acknowledged.
Wiebe, editors. (Thousand Oaks, CA: Sage, 2010), pp. 372-374; Taylor, P. J., G. Catalano, and D.R.F.
Walker. “Exploratory Analysis of the World City Network.” Urban Studies 39 (December 2002): 2377-
2394;Exploratory Research. Wikipedia.
Howell, Martha C. and Walter Prevenier. From Reliable Sources: An Introduction to Historical Methods.
Ithaca, NY: Cornell University Press, 2001; Lundy, Karen Saucier. "Historical Research." In The Sage
Encyclopedia of Qualitative Research Methods. Lisa M. Given, editor. (Thousand Oaks, CA: Sage, 2008),
pp. 396-400; Marius, Richard. and Melvin E. Page. A Short Guide to Writing about History. 9th edition.
IX. Historical Design Boston, MA: Pearson, 2015; Savitt, Ronald. “Historical Research in Marketing.” Journal of Marketing 44
(Autumn, 1980): 52-58; Gall, Meredith. Educational Research: An Introduction. Chapter 16, Historical
Research. 8th ed. Boston, MA: Pearson/Allyn and Bacon, 2007.
a type of observational study sometimes referred to as a analyze their findings. Lack of information can severely
panel study. limit the type of analyzes and conclusions that can be
reached. In addition, the more dissimilarity there is in the
What do these studies tell you? results among individual studies [heterogeneity], the more
difficult it is to justify interpretations that govern a valid
1. Longitudinal data facilitate the analysis of the synopsis of results.
duration of a particular phenomenon.
2. Enables survey researchers to get close to the A meta-analysis needs to fulfill the following requirements
kinds of causal explanations usually attainable to ensure the validity of your findings:
only with experiments.
3. The design permits the measurement of Clearly defined description of objectives,
differences or change in a variable from one including precise definitions of the variables and
period to another [i.e., the description of patterns outcomes that are being evaluated;
of change over time]. A well-reasoned and well-documented
4. Longitudinal studies facilitate the prediction of justification for identification and selection of the
future outcomes based upon earlier factors. studies;
Assessment and explicit acknowledgment of any
What these studies don't tell you? researcher bias in the identification and selection
of those studies;
1. The data collection method may change over Description and evaluation of the degree of
time. heterogeneity among the sample size of studies
2. Maintaining the integrity of the original sample reviewed; and,
can be difficult over an extended period of time. Justification of the techniques used to evaluate
3. It can be difficult to show more than one variable the studies.
at a time.
4. This design often needs qualitative research data What do these studies tell you?
to explain fluctuations in the results.
5. A longitudinal research design assumes present 1. Can be an effective strategy for determining
trends will continue unchanged. gaps in the literature.
6. It can take a long period of time to gather results. 2. Provides a means of reviewing research
7. There is a need to have a large sample size and published about a particular topic over an
accurate sampling to reach representativness. extended period of time and from a variety of
sources.
3. Is useful in clarifying what policy or programmatic
actions can be justified on the basis of analyzing
Anastas, Jeane W. Research Design for Social Work and the Human Services. Chapter 6, Flexible
Methods: Relational and Longitudinal Research. 2nd ed. New York: Columbia University Press, 1999;
research results from multiple studies.
Forgues, Bernard, and Isabelle Vandangeon-Derumez. "Longitudinal Analyses." In Doing Management
Research. Raymond-Alain Thiétart and Samantha Wauchope, editors. (London, England: Sage, 2001),
4. Provides a method for overcoming small sample
pp. 332-351; Kalaian, Sema A. and Rafa M. Kasim. "Longitudinal Studies." In Encyclopedia of Survey sizes in individual studies that previously may
Research Methods. Paul J. Lavrakas, ed. (Thousand Oaks, CA: Sage, 2008), pp. 440-441; Menard, Scott,
editor. Longitudinal Research. Thousand Oaks, CA: Sage, 2002; Ployhart, Robert E. and Robert J. have had little relationship to each other.
Vandenberg. "Longitudinal Research: The Theory, Design, and Analysis of Change.” Journal of
Management 36 (January 2010): 94-120; Longitudinal Study. Wikipedia. 5. Can be used to generate new hypotheses or
highlight research problems for future studies.
XI. Meta-Analysis Design 1. Small violations in defining the criteria used for
Definition and Purpose content analysis can lead to difficult to interpret
and/or meaningless findings.
Meta-analysis is an analytical methodology designed to 2. A large sample size can yield reliable, but not
systematically evaluate and summarize the results from a necessarily valid, results.
number of individual studies, thereby, increasing the 3. A lack of uniformity regarding, for example, the
overall sample size and the ability of the researcher to type of literature reviewed, how methods are
study effects of interest. The purpose is to not simply applied, and how findings are measured within
summarize existing knowledge, but to develop a new the sample of studies you are analyzing, can
understanding of a research problem using synoptic make the process of synthesis difficult to perform.
reasoning. The main objectives of meta-analysis include 4. Depending on the sample size, the process of
analyzing differences in the results among studies and reviewing and synthesizing multiple studies can
increasing the precision by which effects are estimated. be very time consuming.
A well-designed meta-analysis depends upon strict
adherence to the criteria used for selecting studies and
the availability of information in each study to properly
RESEARCH DESIGNS Page 7 of 10
http://libguides.usc.edu/writingguide/researchdesigns
Beck, Lewis W. "The Synoptic Method." The Journal of Philosophy 36 (1939): 337-345; Cooper, Harris, Larry
V. Hedges, and Jeffrey C. Valentine, eds. The Handbook of Research Synthesis and Meta-Analysis. 2nd 7. Produces more complete knowledge and
edition. New York: Russell Sage Foundation, 2009; Guzzo, Richard A., Susan E. Jackson and Raymond A.
Katzell. “Meta-Analysis Analysis.” In Research in Organizational Behavior, Volume 9. (Greenwich, CT: JAI
understanding of the research problem that can
Press, 1987), pp 407-442; Lipsey, Mark W. and David B. Wilson. Practical Meta-Analysis. Thousand Oaks, be used to increase the generalizability of
CA: Sage Publications, 2001; Study Design 101. Meta-Analysis. The Himmelfarb Health Sciences Library,
George Washington University; Timulak, Ladislav. “Qualitative Meta-Analysis.” In The SAGE Handbook of findings applied to theory or practice.
Qualitative Data Analysis. Uwe Flick, editor. (Los Angeles, CA: Sage, 2013), pp. 481-495; Walker, Esteban,
Adrian V. Hernandez, and Micheal W. Kattan. "Meta-Analysis: It's Strengths and Limitations." Cleveland
Clinic Journal of Medicine 75 (June 2008): 431-439.
What these studies don't tell you?
3. While a philosophical analysis may render sample is taken and analyzed. This provides
problematic that which was once simple or opportunities for continuous improvement of
taken-for-granted, the writing can be dense and sampling and methods of analysis.
subject to unnecessary jargon, overstatement,
and/or excessive quotation and documentation. What these studies don't tell you?
4. There are limitations in the use of metaphor as a
vehicle of philosophical analysis.
1. The sampling method is not representative of the
5. There can be analytical difficulties in moving
entire population. The only possibility of
from philosophy to advocacy and between
approaching representativeness is when the
abstract thought and application to the
researcher chooses to use a very large sample
phenomenal world.
size significant enough to represent a significant
portion of the entire population. In this case,
moving on to study a second or more specific
sample can be difficult.
Burton, Dawn. "Part I, Philosophy of the Social Sciences." In Research Training for Social Scientists.
(London, England: Sage, 2000), pp. 1-5; Chapter 4, Research Methodology and Design. Unisa 2. The design cannot be used to create conclusions
Institutional Repository (UnisaIR), University of South Africa; Jarvie, Ian C., and Jesús Zamora-Bonilla,
editors. The SAGE Handbook of the Philosophy of Social Sciences. London: Sage, 2011; Labaree, Robert and interpretations that pertain to an entire
V. and Ross Scimeca. “The Philosophical Problem of Truth in Librarianship.” The Library Quarterly 78
(January 2008): 43-70; Maykut, Pamela S. Beginning Qualitative Research: A Philosophic and Practical
population because the sampling technique is
Guide. Washington, DC: Falmer Press, 1994; McLaughlin, Hugh. "The Philosophy of Social Research." not randomized. Generalizability from findings is,
In Understanding Social Work Research. 2nd edition. (London: SAGE Publications Ltd., 2012), pp. 24-
47; Stanford Encyclopedia of Philosophy. Metaphysics Research Lab, CSLI, Stanford University, 2013. therefore, limited.
3. Difficult to account for and interpret variation
from one sample to another over time,
particularly when using qualitative methods of
XV. Sequential Design
data collection.
Definition and Purpose