PSYB70H3 - Unit 5

Construct Validity

Construct Validity

The extent to which study variables represent the intended theoretical constructs.

Construct

A key concept or attribute being measured.

Assessment Framework

Includes design, reliability, and validity in measuring variables.

Design

Defining, designing, and scaling variables.

Reliability

Consistency across items, raters, or time.

Validity

Accuracy in measuring intended constructs.

Theoretical Construct

The concept researchers seek to understand.

Conceptual Definition

Defines features of a construct and its relationship to other constructs.

Hypothetico-Deductive Method

Observed phenomenon → Theory → Prediction → Testing → Refinement.

Grounded Theory

Theory developed from qualitative data.

Hypothesis Testing

Using theory to make and test predictions.

Operational Definition

Specifies how a variable is measured or manipulated.

Variable

An attribute that varies and is measurable in quantitative research.

Constant

An aspect that does not vary across a study.

Qualitative Research

Analyzes behaviors and imagery for detailed description.

Quantitative Research

Measures variables numerically to test theories.

Self-Report

Participants' own thoughts and feelings.

Behavioral Observation

Recording participants' behaviors.

Physiological Measures

Monitoring bodily processes (e.g., heart rate).

Manipulated Variables

Experimental conditions altered by the researcher.

Response Sets

Consistent but inaccurate responding.

Social Desirability

Faking responses to appear favorable.

Lack of Awareness

Biased by memory or emotions.

Reverse-Worded Items

Detects acquiescence.

Anonymity

Reduces biased responses.

BRUSO

Ensures Brief, Relevant, Unambiguous, Specific, and Objective wording.

Levels of Measurement

Nominal, Ordinal, Interval, Ratio.

Nominal

Categories only.

Ordinal

Rank-ordered.

Interval

Ordered with equal intervals, no true zero.

Ratio

Ordered with equal intervals and a true zero.

Likert-Type Scale

An ordinal scale with ordered responses.

Inventory

Averaged Likert scores treated as interval level.

Interrater Reliability

Consistency across observers.

Test-Retest Reliability

Consistency over time.

Internal Reliability

Consistency across items.

Reliability Coefficients

Interpretation: .90: Strong, .70: Acceptable, .50: Questionable, <.50: Poor.

Validation Study

An empirical study assessing the reliability and validity of measurement tools.

Content Validity

Consistency with theories and research.

Multidimensional Constructs

Defined by multiple dimensions.

Qualitative Approach

Based on theory and expert feedback.

Quantitative Approach

Statistical validation (e.g., factor analysis).