Chapter 5 Flashcards
Operationalizing Variables
a. Self-report: Record people’s answers to questions about themselves or others
i. Questionnaire or interview
ii. Ex: Subjective wellbeing scale
b. Observational: Recording observable behaviors or physical traces of behaviors
c. Physiological: Recording biological data such as hormone levels, heart rate, or change in blood flow in the brain
Scales of Measurement
a. Categorical- Nominal: designate a group, may be text or numeric, higher number does not designate a better group
b. Quantitative: Scales that involve meaningful numeric values that indicate the order and sometimes distance between different levels of a variable [NOIR]
i. Ordinal (Ranks): meaningful order to numeric values, but with unequal or unknown intervals between whole units on the scale
ii. Interval: equal intervals between whole units on the scale, an order to the values, but no meaningful or true zero
iii. Ratio: equal intervals between whole units on the scale, an order to the values, and a meaningful/true zero
1. True zero: The complete absence of a construct (ex: height)
2. Interval and ratio referred to as scale variables
Types of Reliability
a. Reliability: How consistently a construct is measured
i. It is necessary for good construct validity.
b. Test-retest reliability: Consistency in measurement between different testing occasions
i. Expect people with high scores on a test on a variable at time one to also have high scores at time two
ii. Ideal for constructs expected to be stable across time
c. Interrater reliability: the degree to which two or more independent observers will come up with consistent findings
i. Relevant for observational measures
d. Internal reliability: The degree to which a person provides similar ratings or scores for related questions; consistency between items in the same scale
i. Average Inter-Item Correlation
ii. Cronbach’s Alpha: Single number intended to represent the internal reliability
iii. If we ask the same question in different ways, we should get the same answers
Correlation Coefficients
a. R is a numerical value that expresses the strength and direction of a linear association between two variables
b. Varies between -1 and +1 with 0 representing a complete lack of linear relationship
c. Positive correlation: Variables increases/dec together; negative: 1 variable increases, 1 decrease
d. Kappa can be used to measure the extent to which 2 raters place people, entities, or objects into the same categories
Types of Construct Validity
2 subjective. 3 empirical.
5) Face validity: Does the measure seem appropriate?
a. Subjective
6) Content validity: Does it contain appropriate elements?
a. Subjective
b. Ex: contains all aspects of intelligence
7) Criterion validity: examines the extent to which a measure is associated with a behavioral outcome that it should be associated with
a. Also called predictive (future outcomes), or concurrent (present outcomes)
b. Empirical
c. known-groups paradigm: Examine whether scores on a measure distinguish between previously identified groups (people diagnosed with depression score high on BDI)
8) Convergent validity: Examines how well 2 measures of the same constructs correlate with one another (it should)
a. Empirical
9) Discriminant validity: Examines the extent to which measures of dissimilar constructs correlate with one another (it shouldn’t)
a. Empirical
Relationship between reliability and Validity
a. Good reliability is a prerequisite for good validity. However, good reliability does not guarantee good validity.
i. A measure should produce consistent results in order to determine if it is accurate; BUT just because it is consistent doesn’t mean it’s accurate.
ii. Ex: Head circumference to measure IQ is consistently measured but doesn’t accurately measure IQ (reliable but not valid)
