Chapter 3 The Measurement of Behavior Continued Validity

Chapter 3: The Measurement of Behavior Continued Validity: the extent to which a measure actually measures what it is intended to measure • The truthfulness of a measures. • A test can be reliable and not be valid. Construct validity: The extend to which a measures the conceptual variable it is designed to measure. – Related to the adequacy of the operational definition of the variable

• Convergent validity: the extent to which scores on a measure correlate with scores on a different measure of the same construct. – • Discriminant validity: the extent to which scores on a measure do NOT correlate with scores on a measure of a different construct. – • Does depression measure A correlate with depression measure B? Does the depression measure correlate with verbal ability? Content validity: whether the measure covers the full domain of the conceptual variable. – Does your test measures all the variables that are involved in IQ?

Face validity: whether a measure appears to measures a construst • Face validity does not mean measure is valid – • If a measure lacks face validity it still may be valid – • measuring skull size appeared to be a good measure of IQ. some personality measures have questions that do not appear to be related to personailty Sometimes researchers do not want face validity. – Some measures of psychotic behavior may have disguised questions so the participant does not know what it is being measured.

1. Criterion Validity: to extent to which scores on a measure are related to a criterion (an indicator of the construct). 1. Does the measure allow you to distinguish participants on a behavioral criterion? 2. The extend to which the measures correlates with a behavioral measures. 2. Predictive validity: the extent to which a measure correlates with a future behavior. 1. Does the GRE predict success in grad school? 3. Concurrent validity: whether a measure correlates with a behavioral measures at the same time.

Validity Issues • Reactivity: a measure is reactive if awareness of being measured influences the participant’s behavior and scores. • Test-bias: occurs when a measure is not equally valid for everyone who takes the test. – Difficult to determine because it is hard to detect whether the groups actually do differ on the conceptual variable. – A biased test will have a weaker predictive validity for the group which it is thought to be biased against.

Ways to improve validity of measures: • Pilot testing: try out the measure on a small group of participants before the study starts to get an idea of how they react to it and whether it is measuring what it is intended to measure. • Make your items non-reactive • Consider face and content validity by choosing items that seem reasonable and represent a broad range of questions concerning the concept. • Validate your measure against other measures and criterion variables

Chapter 4: Approaches to Psychological Measurement Observational Measures • Direct observation of behavior. Make observations of behavior and record them objectively. 1. Naturalistic vs Contrived setting Naturalistic observation: observation of behavior in natural setting with no intrusion of researcher. • Observe everyday behavior and activities • Participant observation: researcher participates in study, usually unknown to the participant.

Contrived Setting: observe behavior in a setting designed for the study (e. g. lab) – observe child aggression in playroom in a lab Ecological validity: extent to which research is conducted in situations that are similar to participants’ everyday lives. 2. Disguised vs Nondisguised Observation Undisguised: participants know that a researcher is observing them • Reactivity: participants may act differently if they know they are being watched.

Disguised: participants do not know they are being observed. • Informed consent? Usually not needed if watching people in public and the researchers do not intrude on or upset participants. • Partial concealment: participants know they are begin observed, but do not know what the researchers are measuring. • Knowledgeable informants: have someone who knows the participants well (friends, family) observe their behavior.

• • Unobtrusive measures: measures taken without the participant knowing it. Dyadic interaction paradigm: Observe behavior naturally without participant knowing and then inform them after and ask for their consent. 3. Behavioral Recording • how to record the participants behavior Narratives: record a full description of participant’s behavior. Unstructured. – – Video or audio recording field notes

Checklists: Structured. Record behavior by checking off behaviors on a checklist. • Depends on the operational definition of the concept – some behaviors may be difficult to record on a checklist – have to know what you are looking for ahead of time Temporal measures: when behavior occurs (latency) or who long it lasted (duration). • Reaction time: time between stimulus and response.

1. Task completion time: time to complete a task. 2. Interbehavior latency: time between two behaviors. 3. Duration: how long a behavior lasted. Observational Rating Scale: measure quality or intensity of behavior. 1. Scale to measure intensity of laughter

Physiological Measures • Measure of neural activity – • Measure of autonomic nervous system activity – • Involuntary responses like heart rate, temperature, and respiration. Blood analysis – • EEG to measure brain waves hormones, blood cells Specific bodily reactions – arousal, blushing.

Self-Report: Questionnaires and Interviews • Survey research: ask people to provide information about themselves. • Most widely used method of collecting descriptive information on participants. • Goal is to produce a ‘snapshot’ of the opinions, attitudes, or behaviors of a group of people at a given time. • Allows researcher to gather a lot of information in a short amount of time.

Questionnaires: a set of fixed format, self-report items completed by the participant. – Can be administered individually or to a group. – Can be mail-out surveys or internet surveys. Response rate: percent of people who complete and return and questionnaire. • Important to have a high response rate to ensure your results reflect the responses of the entire group (rather than a subset) to ensure your results are generalizabile.

Interviews: questions are read to the participant in person or over the phone and the participant responds orally. • Unstructured interview: the interviewer talks freely with the participant about many topics. – Provides in-depth information, but may be subjective. • Structured interview: interviewer uses quantitative fixed-format items. Questions are prepared ahead of time. – allows better comparison across individuals. • Focus group interview: interview a selected groups on individuals at the same time.

Pros and Cons of Questionnaires and Interviews: • Questionnaires are cheaper, less time consuming, and require less training than interviews. • Interview bias: interviewers can bias participants’ answers by their facial expressions and intonation. • Questionnaires can produce more honest answers, especially when dealing with sensitive issues. • Questionnaires are inappropriate for those who can not read. • Interviews are better for young children and cognitively impaired or disturbed individuals. • Interviewers can ensure the participants understand the questions

Biases in self-report measures Social Desirability: people may respond in the socially desirable manner. • They may be concerned that they may be judged by the researcher for having a particular opinion. • Threats the validity of the measure. – Ensure questions are worded neutrally – Ensure anonymity of responses Response styles • Acquiescence: tendency to agree with statements • Nay-saying: tendency to disagree with statements.

Archival Data • Analyze data from existing records – databases, census, magazines, newspapers • Useful for studying psychological phenomena that occurred in the past – changes in family types • Useful for studying behavioral changes over time • Allows the researchers to study phenomena that they cannot conduct a study on – riots, world events.