4 terms

Four Characteristics of Good Psychological Research

psychological studies vary tremendously in design, but most good research shares these four attributes.
1. Theoretical Framework
Theory: a systematic way of organizing & explaining observations. Provides the framework for generating a hypothesis: proposed cause-and-effect relationship between 2 or more variables, that flows from the theory or an important question.

Variable: a phenomenon that changes across circumstances or varies among individuals. Can be-

continuous variable: can be placed on a continuum of infinite values (intelligence, body weight)

categorical variable: can take on fixed values (sex- male or female, race, had heart attack-yes or no)
2. Standardized Procedures
Expose the participants in a study to procedures that are as similar as possible.
Procedure that is the same for all subjects except where variation is introduced to test a hypothesis.

ex. Pennebaker study of emotional expression & health, the experimenters instructed students in both groups to write for 20 minutes a day for 3 days.
3. Generalizability of Research
Sample that is representative of the population as a whole
-researchers take samples from a limited portion of the entire population so sampling must be representative as a whole so that conclusions drawn from the samples are likely to be true of the rest of population

Procedure that is sensible and relevant to circumstances outside the laboratory
-for a study to be generalizable its procedure must be valid:
Internal Validity (valid design- are methods/procedures of study sound or flawed?)
External Validity (does experimental situation resemble the situation found in the real world, outside the lab?)
Leads to Experimenter's Dilemma because there is a trade-off in which researchers must choose to place more emphasis on external validity or on internal validity.

Generalizability: the applicability of the findings to the entire population of interest to the researcher.
4. Objective Measurement
Measures that are reliable (that produce consistent results)
1. Test-retest reliability (consistency across time)
2. Inter-rater reliability (consistency across people)
3. Alternate form (Least popular-consistency across forms)
4. Internal consistency (Most popular-consistency across items)

Measures that are valid (that assess the dimensions they purport to assess)
1. Face validity (Least important-whether or not it appears valid by looking at it)
2. Content validity (does content measure a well-defined body of material?)
3. Criterion validity (establishing the relationship between test performance and some external criterion- concurrent & predictive)
4. Construct validity (Most important-measures what if claims to measure-convergent & discriminant)

Construct: variable we want to measure (e.g., intelligence)

Measure: a concrete way of assessing or operationalizing an abstract construct.
-e.g., Pennebaker study, variable "health" was operationalized as the # of times a student visited the health clinic.

To study a construct, a researcher must first devise a technique to measure it.

Researchers must devise ways to quantify or categorize variables so they can be measured objectively.