Psych 311 Unit 2 study guide

Helpfulness: 0
Set Details Share
created 5 years ago by wouhib
updated 5 years ago by wouhib
show moreless
Page to share:
Embed this setcancel
code changes based on your size selection

Validity: definition, types, difference between types, ability to determine which types of validity evidence would be most salient/important for different types of research questions

Definition: Does the instrument measure what it is supposed to measure?

- Face: Does the instrument appear to measure the construct that it claims to measure. Example: An instrument assessing sleep includes questions about sleep.

- Concurrent: Scores from one instrument are related to scores from another instrument measuring the same variable, given at the same time. Example: A new instrument measuring intelligence classifies students in the same way as standardized IQ tests given in the same time period

- Predictive: Scores on an instrument accurately predict behavior or a score in the future. Example: SAT scores predict first year college GPA

- Construct: Measurements of a variable behave in the same way as the variable itself. Example: Feminists describe themselves as endorsing specific beliefs that are also endorsed on your measure of feminism

- Convergent: Strong relationship between the scores of two different measures used to measure the same construct. Example: Teacher ratings and observational scores for child aggression in schools

- Divergent: Little or no relationship between the scores of two different constructs when measured with the same method. Example: Teacher and observational scores for child activity level and aggression are separate and distinct


Reliability: definition, role of error, sources of error, systematic vs random error, be able to explain how the number of items in a scale impacts reliability

Definition: Does the instrument measure something consistently?

Sources of error: observer error, environmental changes, participant changes

Random error: Caused by unpredictable fluctuations & inconsistencies in the measurement process (e.g., earthquake, fire alarm, spider dropping onto your desk)

Systematic error: Typically constant or proportionate to what is presumed to be the true value of the variable being measured (e.g., your scale is off by 2 pounds, but it is always off by 2 pounds)

As the number of items in a scale increases, reliability also increases


Types of reliability


- Test-retest: A type of successive reliability in which the researcher/evaluator compares scores from two measurements of the same person

- Parallel forms: A subtype of test-retest reliability in which two versions of the same test are given to the same person at different times to see if the measurements are (enough) the same


- Inter-rater: Degree of agreement between two observers who simultaneously record measurements of participant behaviors

- Split-half: A measure of internal consistency, Splitting a questionnaire in half (not necessarily 1st half and second half), computing a separate score for each half, and determining the correlation of the two scores


interpretation of reliability scores, questions about example reliability scores

Correlation coefficients: A numerical value that measures or describes the relation between two variables, typically represented by r

- The direction of relation: +/-

- The form of relation: indicated by type (pearson or spearman)

- Strength or consistency of relation: -1 to +1, 0=no relation

- Degree of relation r2

- Small: r=.1 or r2=.01

- Medium: r=.3 or r2=.09

- Large: r=.5 or r2=.25


Relationship between reliability and validity

card image

Reliability is simple therefore it is reported more often, also if measure is not reliable then it cannot be valid


Find, identify, interpret information about psychometrics of measurement instruments in the method section of an empirical article



Experimenter bias

When measurements obtained in a study are influenced by the experimenters expectations or personal beliefs regarding the outcome of the study

Single-blind: PARTICIPANT does not know predicted outcome (note this is different from your textbook information)

Double-blind: Researchers and participants are BOTH unaware of predicted outcome


Demand characteristics

Features of the study that suggest to the participant what the purpose and hypothesis of the study is and influence the participants to respond a certain way


Reactivity and reactivity reduction

Reactivity occurs when participants modify their behavior in response to the fact that they are being measured

To reduce reactivity:

- Observe without participant awareness

- Disguise questions

- Use deception

- Reassure participants that answers are confidential and anonymous


Hawthorne studies

Attempt to study what conditions in a factory modified worker behavior but ALL study variables improved worker productivity, because the process of being observed made workers more aware of their work


Subject roles

- Faithful: Follow instructions, don’t modify responses in any way based on hypothesis of what the study is about

- Apprehensive: Socially desirable responses, want to look good and give responses they think will make a good impression

- Negativistic: Respond in a way that they think will specifically contradict the researchers’ hypothesis

- Good: Trying to be helpful, respond in a way that they think will most help the researcher get the data they need/want