Research Methodology and Data Analysis Key Concepts

Flashcards covering key research concepts from Chapters 1 through 5, including operational definitions, literature review synthesis, experimental design, scale validity, and inferential statistics.

How are scope and delimitation defined in a research study?

Scope defines the precise operational boundary and demographic or geographical limits within which the study operates, while delimitation explicitly cites what is excluded from the study by choice of the researcher.

What is a Null Hypothesis ($H_0$)?

A statement predicting no statistically significant relationship, difference, or effect between variables, serving as the baseline assumption until proven otherwise.

What is the function of a mediating variable in a study?

It transmits the effect of an independent variable (IV) to a dependent variable (DV), explaining the underlying process, mechanism, or pathway of the relationship to answer 'HOW' or 'WHY' the IV affects the DV.

How does an operational definition of terms differ from a conceptual definition?

A conceptual definition provides general theoretical or dictionary definitions, whereas an operational definition specifies exactly how variables are measured, manipulated, or observed within the specific context of the study.

What is the key difference between a theoretical framework and a conceptual framework?

A theoretical framework relies on established, formally tested theories from literature, while a conceptual framework is the researcher's specific operational model and representation of variable interactions.

What is a research gap and what role does it play in Chapter 1?

A research gap pinpoints unanswered questions, methodological limitations, or unresolved contradictions in existing literature, serving as the core justification for conducting the study.

What defines a methodological literature review?

It is a synthesis technique where past studies are grouped based on shared research designs, methodologies, paradigms, or data collection tools, focusing on how studies were conducted rather than just what they found.

Why is literature synthesis preferred over sequential summaries in a review of related literature?

Sequential summaries merely list and rephrase individual papers, whereas literature synthesis integrates empirical findings across studies to reveal overarching trends, gaps, and contradictions.

How should researchers reconcile contradicting empirical findings across different years (e.g., 2018 vs. 2023)?

Researchers should analyze contextual evolution, such as technological advances or infrastructure changes, rather than ignoring older data.

What analytical flaw occurs when synthesizing literature without critical evaluation?

Uncritically weighing low-rigor, high-bias studies (e.g., small $N$, no control) equally with robust designs is a severe analytical flaw, so source quality and rigor must be appraised.

What are the essential requirements of a True Experimental Design?

It requires manipulating an independent variable while randomly assigning participants to control and experimental groups to establish direct cause-and-effect.

What characterizes an Ex-Post Facto Research Design?

It is an investigation where the independent variable has already occurred naturally and cannot be manipulated or randomly assigned, meaning direct causality cannot be conclusively established.

How is Stratified Random Sampling executed?

By dividing the target population into non-overlapping homogeneous subgroups (strata) and taking a random sample from each subgroup proportional to its size.

When is Snowball / Chain-Referral Sampling most appropriately used?

It is ideal for hard-to-reach, rare, hidden, or sensitive populations (such as survivors of a rare institutional crisis), where existing participants recruit future subjects.

What does Content Validity evaluate in a measurement instrument?

It ensures that a survey instrument comprehensively covers all dimensions, domains, and aspects of the theoretical construct being measured.

What does Cronbach's Alpha measure, and what threshold indicates acceptable scale reliability?

It measures internal consistency for Likert-scale items—how reliably and closely related items are in measuring a single construct; a value of Cronbach's alpha $> 0.70$ generally indicates acceptable reliability.

What is the primary purpose of conducting a pilot study before main data collection?

To serve as a dry run that assesses item clarity, factor structure, and internal consistency reliability in order to refine instruments before full-scale deployment.

How does triangulation enhance qualitative trustworthiness?

By cross-verifying qualitative findings across multiple data sources (such as interviews, observations, and documents) to establish convergent patterns and construct credibility.

What is Bracketing / Epoché in phenomenological research?

The practice where a researcher consciously sets aside personal biases, assumptions, and presuppositions to understand participants' lived experiences objectively.

What sampling bias and ethical flaws arise from using authority figures (e.g., CEOs) for data collection?

It induces coercion and social desirability bias, leading to severe selection and sampling bias.

What do the Mean and Standard Deviation ($SD$) measure in normally distributed continuous data?

The mean represents central location, while the standard deviation ($SD$) represents score dispersion around the mean, where a large $SD$ indicates high heterogeneity.

What occurs during the Axial Coding stage of qualitative analysis?

Initial open codes are grouped into broader, higher-level conceptual categories and sub-themes to connect codes into themes.

What does the Coefficient of Determination ($R^2$) represent in regression analysis?

It indicates the proportion of total variance in the dependent variable that is predictable from the independent variable (e.g., $R^2 = 0.45$ means $45\%$ of $DV$ variance is explained).

What is the decision rule for hypothesis testing when $p < \alpha$ (e.g., $p = 0.03$ vs $\alpha = 0.05$)?

Reject $H_0$ because the probability of obtaining such results by chance under the null hypothesis is less than $5\%$ ($p < 0.05$), signifying statistical significance.

Why does statistical significance ($p < 0.05$) not necessarily imply practical significance?

A very large sample size ($N$) can produce a statistically significant $p$-value even for a tiny difference; effect size must be evaluated to measure real-world magnitude.

Why does a strong positive correlation (e.g., Pearson $r = 0.85$) fail to prove direct causality?

Correlation indicates association, not direct causation, because it ignores potential confounding third variables and reverse causality.

How are statistical tests selected based on variable types and study conditions?

Two-Way Factorial ANOVA analyzes 2 categorical IVs on 1 continuous DV; ANCOVA controls for continuous baseline covariates statistically; Chi-Square tests association between 2 categorical variables; Fisher's Exact / Category Collapsing is used when $> 20\%$ of expected cell frequencies in Chi-Square are $< 5$.

How should conflicting quantitative and qualitative findings be handled in Divergent Integrated Analysis?

Researchers must analyze underlying paradoxes rather than suppressing data, because conflicting data reveals complex, multi-faceted realities.