Quantitative Research Methods Flashcards

Which nonparametric test is the appropriate alternative to a one-sample t-test when the normality assumption is violated?

Answer: Wilcoxon signed-rank test

The Wilcoxon signed-rank test is the nonparametric equivalent of the one-sample (or paired) t-test when data are not normally distributed.

A researcher wants to predict a binary outcome (pass/fail) from several continuous predictors. The most appropriate analysis is:

Answer: Logistic regression

Logistic regression models the probability of a binary outcome as a function of predictor variables, producing odds ratios.

The standard error of the mean (SEM) decreases as:

Answer: Sample size increases

SEM = σ/√n, so increasing sample size reduces the SEM, making the sample mean a more precise estimate of the population mean.

Content validity is best assessed by:

Answer: Having subject matter experts review whether items cover all relevant domains

Content validity is typically evaluated through systematic review by domain experts to confirm comprehensive coverage of the construct.

In a normal distribution, approximately what percentage of data falls within two standard deviations of the mean?

Answer: 95%

The empirical rule states that approximately 95% of data in a normal distribution lies within ±2 standard deviations of the mean.

Which experimental design is most effective for controlling for individual differences by having each participant experience all conditions?

Answer: Within-subjects (repeated measures) design

Within-subjects designs use each participant as their own control, eliminating between-person variability from the error term.

When a measurement instrument consistently produces scores that are 5 points too high, this is an example of:

Answer: Systematic error (bias)

Systematic error introduces a consistent, directional offset in measurements, which affects validity but not necessarily reliability.