Designing a Study You Can Defend
Online focus groups can generate credible qualitative evidence, but comparability with in-person discussions depends on interaction, participant access, communication mode, and analytical purpose. Learn when online focus groups are appropriate, what differences matter, and how to evaluate data quality without assuming that either mode is inherently superior.
Read Guide →
Designing a Study You Can Defend
Participant and nonparticipant observation differ in how researchers engage with the activities and people they study. Learn when participation improves contextual understanding, when observational distance is preferable, and how researcher roles affect ethics, data quality, and interpretation.
Read Guide →
Designing a Study You Can Defend
Structured observation uses predefined categories to record selected behaviors, while unstructured observation allows researchers to investigate events and interactions more openly. Learn when each approach is appropriate, how their evidence differs, and whether both can be combined within one study.
Read Guide →
Designing a Study You Can Defend
Overt observation makes the researcher's observational role known, while covert observation involves concealing that role from some or all participants. Learn how this distinction affects consent, participant behavior, access, research ethics, data interpretation, and the circumstances under which covert observation may be defensible.
Read Guide →
Designing a Study You Can Defend
Choosing what to observe is a consequential research design decision. Learn how to translate research questions into observable behaviors and events, establish recording boundaries, avoid misleading indicators, and determine when exploratory observation is needed before developing a structured protocol.
Read Guide →
Designing a Study You Can Defend
Observational research rarely captures every relevant behavior across all times and settings. Learn how to select observation periods, locations, events, and interactions systematically, distinguish time sampling from event sampling, and ensure that the resulting evidence supports your research question.
Read Guide →
Designing a Study You Can Defend
Participants may change their behavior when they know they are being observed, but researcher presence does not always produce predictable effects. Learn how observation reactivity differs from the Hawthorne effect, when these influences matter, and how to evaluate their implications for observational research.
Read Guide →
Designing a Study You Can Defend
Using multiple observers may improve coverage, reveal differences in interpretation, and support assessments of coding consistency. Learn when additional observers are justified, how to distinguish independent observation from collaborative interpretation, and why high interobserver agreement does not automatically establish validity.
Read Guide →
Designing a Study You Can Defend
Documents and existing texts can provide the principal evidence for a research study without requiring interviews, surveys, or observations. Learn when documentary data are sufficient, how to evaluate their origins and limitations, and how to distinguish documents as primary evidence from documents used merely as background literature.
Read Guide →
Designing a Study You Can Defend
Research documents extend beyond printed reports and official policies. Emails, digital records, websites, meeting minutes, photographs, and other recorded materials may qualify depending on the research question. Learn how to define a document, distinguish documentary evidence from other data, and establish defensible boundaries for a research corpus.
Read Guide →
Designing a Study You Can Defend
A document corpus is more than a collection of available files. Learn how to establish its boundaries, define eligible documents, identify sources, manage versions and duplicates, document retrieval decisions, and determine whether the resulting corpus can support your research questions.
Read Guide →
Designing a Study You Can Defend
Should researchers analyze every available document or select a sample? Learn how corpus size, research objectives, analytical depth, representativeness, and resource constraints affect the decision, and how to justify either approach without confusing complete analysis with complete coverage of the documentary population.
Read Guide →
Designing a Study You Can Defend
Public social media posts can provide documentary evidence about communication, institutional representations, and online discourse. However, public accessibility does not automatically establish unrestricted research use. Learn how to evaluate social media content as documentary data, establish a defensible corpus, address consent and privacy, and interpret findings without confusing platform activity with wider public opinion.
Read Guide →
Designing a Study You Can Defend
Quantitative content analysis measures defined features of communication, while qualitative content analysis examines meaning, context, and patterns of interpretation. Mixed-methods content analysis integrates numerical and qualitative evidence for a common research purpose. Learn how to choose the appropriate approach, develop coding procedures, and avoid confusing numerical counts with meaningful interpretation.
Read Guide →
Designing a Study You Can Defend
Document analysis does not automatically require interviews, surveys, or observations. Additional data sources become important when documents cannot adequately establish the behaviors, experiences, implementation processes, or outcomes central to the research question. Learn when triangulation is justified, how to integrate evidence, and why agreement across sources is not always necessary or sufficient for credible findings.
Read Guide →
Designing a Study You Can Defend
Secondary data analysis may be preferable when existing datasets provide suitable measurements, adequate population coverage, and stronger evidence than a feasible new study could collect. Learn how to evaluate data quality, costs, accessibility, analytical limitations, and research objectives before deciding whether to reuse existing information or collect new data.
Read Guide →
Designing a Study You Can Defend
A dataset may contain relevant variables yet remain poorly suited to a new research question because it was originally created for another purpose. Learn how to evaluate conceptual alignment, measurement, population coverage, timing, and data-generation procedures, and how to decide whether the question should be retained, refined, or abandoned.
Read Guide →
Designing a Study You Can Defend
Secondary datasets can support research questions their original investigators never anticipated. However, new analytical possibilities do not automatically establish valid new interpretations. Learn how to assess construct alignment, temporal structure, confounding, sampling, and the distinction between exploratory discovery and confirmatory analysis.
Read Guide →
Designing a Study You Can Defend
Secondary datasets often contain indirect indicators rather than direct measures of the constructs researchers want to investigate. Learn when proxy variables are appropriate, how to justify their relationship to the intended construct, assess measurement limitations, and prevent misleading interpretations.
Read Guide →
Designing a Study You Can Defend
Missing variables can undermine causal inference even when a secondary dataset contains thousands of observations. Learn how to identify essential confounders, distinguish unmeasured variables from missing values, evaluate adjustment strategies, and determine when residual confounding makes a proposed analysis unsuitable.
Read Guide →
Designing a Study You Can Defend
Existing datasets rarely contain every variable a researcher would ideally collect. Learn when refining a research question is methodologically defensible, how to distinguish conceptual improvement from result-driven changes, and when the available data are too limited to justify adaptation.
Read Guide →
Designing a Study You Can Defend
Not every existing dataset can support a defensible secondary analysis. Learn how to distinguish manageable limitations from fundamental design problems, evaluate whether the intended research question remains answerable, and decide when to revise or abandon a proposed study.
Read Guide →
Designing a Study You Can Defend
Medical records, administrative databases, and other routinely collected data can provide valuable research evidence, but their original purpose may differ from the needs of a scientific investigation. Learn how to evaluate their suitability, measurement quality, completeness, population coverage, and ethical use.
Read Guide →
Designing a Study You Can Defend
Incomplete routine records can distort research findings even when datasets contain thousands of observations. Learn how to distinguish missing information from unrecorded events, investigate documentation patterns, and select defensible analytical responses.
Read Guide →
Designing a Study You Can Defend
Record linkage can connect information about the same individuals, events, or organizations across datasets. Learn when linkage is methodologically necessary, how to assess whether it is feasible, and when its risks may outweigh its benefits.
Read Guide →