W7 - Research Methods: Surveys and Experiments

Comprehensive practice questions covering survey methodology, experimental design, validity types, sampling, and statistical interpretation based on lecture notes.

What are the five elements of a true experiment?

Manipulation, measurement, comparison, and control, random assignment.

How does an experiment establish chronology by design?

By manipulating the variable before measuring the outcome.

What is the difference between internal and external validity?

Internal validity asks if the relationship is real and caused by what you think; external validity asks if findings hold outside the specific study (different people, settings, and times).

What is 'acquiescence' in respondent characteristics?

When people agree to everything said regardless of whether it reflects their actual views.

What are the four steps in the 4 Step Response Model?

1. Comprehension, 2. Retrieval, 3. Judgement and Estimation, 4. Reporting.

How can researchers detect question sensitivity?

Through high non-response rates.

What is the Randomized Response Technique?

A technique where a random event (like a coin flip) determines if a participant answers truthfully or gives a forced answer, allowing group rates to be calculated while individual answers remain hidden.

What is the Item Count Technique?

A method where respondents report only the count of items applicable to them from a list, with a sensitive item hidden among harmless ones.

Define a 'Double Barreled' question.

A question that includes two questions in one.

Distinguish between Primacy and Recency effects in response order.

Primacy is the tendency to select the first option (common in questionnaires); Recency is the tendency to select the last options as they are easier to recall (common in interviewer-led surveys).

Contrast the 'Contrast Effect' and the 'Assimilation Effect' in questionnaires.

The contrast effect makes the current question seem more extreme because of a preceding one; the assimilation effect pulls the current answer in the same direction as the preceding question.

What are the four computer-assisted modes of survey data collection?

CAPI (Personal Interview), CATI (Telephone Interview), CASI (Self-Interview), and CAWI (Web Interview).

What is a 'Systematic Sample'?

Selecting every $k^{th}$ unit from a sampling frame.

Name the four types of error in surveys.

Coverage Error, Sampling Error, Non-response Error, and Measurement Error.

What is 'Cognitive Interviewing' in pre-testing?

A small sample of respondents completes a survey while thinking aloud to reveal comprehension and retrieval problems.

What are the three requirements for establishing causality?

1. Covariation, 2. Chronological Order ($X$ precedes $Y$), 3. No Alternative Explanations.

What is the 'Cross-lagged Panel Model' used for?

Testing temporal order in panel surveys to see if variable $X_{t1} \rightarrow Y_{t2}$ and/or $Y_{t1} \rightarrow X_{t2}$ to suggest a causal direction.

What sample type is represented by the acronym WEIRD?

Western Educated Industrialized Rich Democratic.

What is the difference between Mundane Realism and Experimental Realism?

Mundane realism is whether the setting looks like real life; experimental realism is whether the participant feels they are in a real situation.

Distinguish between Noise and Bias in experiments.

Noise (Random Error) involves individual differences that vary randomly, increasing within-group variance; Bias (Systematic Error) involves individual differences systematically distributed between conditions, creating a confound.

Define 'Resentful Demoralization' and 'Compensatory Rivalry'.

Resentful demoralization occurs when a control group performs worse because they resent not getting treatment; compensatory rivalry is when a control group works harder to prove the treatment doesn't matter.

When does an interaction effect exist?

When the effect of one factor depends on the level of another factor.

What are the three types of manipulation checks?

Participation Check (encountering the manipulation), Attention Check (following instructions), and Perception Check (perceiving the manipulation as intended).

What is 'Counterbalancing' in within-subjects designs?

Varying the order of conditions (e.g., Group 1 does $A \rightarrow B$, Group 2 does $B \rightarrow A$) to distribute order effects equally.

Define 'Regression to the Mean'.

The tendency for extreme scores at time 1 to move toward the average at time 2 due to measurement error.

What is a 'Mixed Design'?

A design that uses both within-subject and between-subject factors.

In an ANOVA, what do 'Between Group Variance' and 'Within Group Variance' represent?

Between Group Variance is the 'Signal' (differences between group averages); Within Group Variance is the 'Noise' (spread of people inside each group).

When interpreting SPSS 'Sig.' values, what threshold indicates significance?

$Sig. < 0.05$ is significant; $Sig. > 0.05$ is not significant.

Which two checks in survey/experiment analysis specifically desire a non-significant result ($Sig. > 0.05$)?

Levene’s test and the Randomization check.