UCLA Psych 100B Final | Quizlet

Operational Definition

turning abstract concepts into things we can manipulate and/or measure

construct validity

how well does our chosen operationalization of a variable represent the construct we intend to study

"independent variable" for correlational studies

predictor variable

"dependent variable" for correlational studies

Criterion variable

Four different kinds of scale

- nominal

- ordinal

- interval

- ratio

Nominal Scale

- labels

- you just know that categories are different/distinguishable from each other but you don't know much about it

- there is no intrinsic ordering to the categories (no agreed way to order them from highest to lowest)

examples of nominal variables

Gender - Male, female

Hair color - blonde, brown, brunette, red

Property - houses, condos, townhouses

States - California, NY, Penn

Cup size example -

◦ Predator

◦ King Kong

◦ Godzilla

ordinal variable

- Categorical but has a clear ordering of variable

Even though we can order these from lowest to highest, the spacing between the values may not be the same across the levels of the variables.

examples of ordinal variable

- economic status - low, medium, high

- Educational experience - high school grad, some college, college grad

note the unequal spacing: The difference between categories one and two (elementary and high school) is probably much bigger than the difference between categories two and three (high school and some college)

Cup size example

◦ Godzilla (largest)

◦ King Kong (medium)

◦ Predator (smallest)

Interval Variable

- similar to ordinal variable but the categories are equally spaced

Example of interval variable

- annual income that is measured in dollars; we have three people who make $10,000, $15,000 and $20,000. The second person makes $5,000 more than the first person and $5,000 less than the third person, and the size of these intervals is the same.

- Cup size example

◦ Predator

◦ King Kong (4 oz. more than Predator)

◦ Godzilla (4 oz. more than King Kong)

* You know how much more you get with Godzilla than with

King Kong, but you still don't know the absolute amount*

Ratio variable

Ratio variables are interval variables, but with the added condition that 0 (zero) of the measurement indicates that there is none of that variable.

Degrees Celsius is ratio variable. True or False?

False

- 0 on ratio scale has to mean there is none of that variable but 0C does not mean that there is no temperature

Other examples of ratio variables?

Height, mass, distance

Reliability has to do with?

Are your measures consistent & Stable

High Replicability means

If we ran this experiment at another university or in another class,

we would still get the same results

high test-retest reliability

If we design a test of agreeableness, and have someone take it

twice, two weeks apart, they should score similarly if our test has good test-retest reliability.

inter-rater reliability

If our raters have good inter-rater reliability, their scores will be highly

correlated

Internal validity

How certain are we that the change in the IV is solely causing the changes in the DV?

confounding variable

An extraneous variable that is varying systematically right along

with our IV

Extraneous variables vary _________; confounding variables vary __________

Extraneous variables vary randomly; confounding variables vary systematically

Experimenter bias

Biases introduced by the experimenter

participant bias

Biases introduced by the participant

Materials bias

Biases introduced by the materials

Specific Item Effects

It may have been the specific items used in your materials (and not your IV manipulation) that caused the differences observed between

conditions

Selection Effects

Participants were not randomly assigned to groups, resulting in differences in the kinds of participants in each group

ex. Letting participants choose the condition they participate in

Carryover/Order Effects

Patterns emerge in our results that are due to the order in which conditions were completed rather than the manipulation.

Practice effects

Participants perform better in one condition just because they completed it last and therefore were more practiced at the task

Fatigue effects

Participants perform worse in one condition just because they completed it last and were tired/fatigued by the task

External validity

- How well does the experiment map onto the real world?

- Can you generalize the results to real world environments?

- Can you generalize the results to other populations?

Factorial design

Two or more IVs

Main effects

Overall effect of an IV

Look at: Marginal means

Interaction effects

Test whether the effect of each IV depends on the level of the

other

Simple effects

Test the effect of one IV at a particular level of another IV

Look at: Cell means

Parallel lines mean _________

Non-parallel lines mean ________

Parallel lines = no interaction

Non-parallel lines = interaction

independent samples t-test is used for ___________ design

between-subjects design

Paired/dependent samples t-test is used for _________ design

Within-subjects design

What to use for single IV with 3+ level

ANOVA

characteristics of quasi experimental designs

Researchers cannot manipulate an IV (cannot randomly assign)

Types of Quasi-Experimental Designs

= Interrupted Time Series

- Pretest - Posttest

- Ex Post Facto (Prospective & Retrospective)

- Developmental (Cross-sectional & Longitudinal)

pre-test post-test design

Measure once before event > event > measure once after

Pre test post test with non equivalent control group

Measure once before event > event > measure once after

+ Compare with a group that hasn't been exposed to event

interrupted time series design

Like pre-test post-test except that you take MANY

measurements both before and after event at regular intervals

Ex Post Facto - Prospective

- looking forward

Selects pre existing groups who have some

characteristic or behavior and follows them forward to

observe potential impacts

Ex Post Facto - Retrospective

- looking backward

Identifies differences among groups and looks

backwards at their lives to identify potential

contributing factors

cross-sectional study

a study in which people of different ages are compared with one another

- between sub

longitudinal study

a study that observes the same participants at different ages over a long period of time

- within-sub design

Facebook is interested in whether a new advertising system can cause people to leave Facebook. Since it has data on the number of

people who quit Facebook everyday, it plans to compare the quitting rate before and after the change. Data is collected once a day for the 5 weeks before the new system, and once a day for the 5 weeks afterwards.

interrupted quasi

Prof. Fowler was interested in the "Colbert Bump" claim. He looked at changes in donations for politicians before and after they appeared on the show. He looks at donation data at a single time point before they appear and after they appear. He compares this to a

group of matched politicians who do not appear on the show.

pre test post test + non equivalent control group

Dr. Reynolds was interested in determining whether students with a higher GPA suffered from less depression. He asked students: 1) what is your GPA and 2) rate your level of depression on an interval scale

(0 = not at all depressed, 10 = severely depressed)

Correlational

History effects

Real world events may have occurred that caused a change in the

thoughts, feelings, and behavior of your participants

Testing Effects

Simply having taken the test before causes people to perform better the second time

Maturation effects

Subjects mature naturally over time, which causes some changes in how they think, feel, and behave

Selection effects

Because you couldn't randomly assign subjects to groups, perhaps your subjects were different at the outset.

Statistical Regression to the Mean:

when you have people perform at extremes, the next time you

measure them, they are more likely to be closer to their mean

Cohort effect

In a cross-sectional design, differences observed between groups may not simply be due to the difference in ages, but other much more complex generational differences

Mortality effect

- some subjects drop out of your study

- Some conditions (the more difficult, more demanding conditions)

may experience higher dropout rates

- Ex. people who are bad at math drop out + mean test scores become higher because we have fewer ppl who are bad at math

Ways of Quantifying Observations

- Frequency

- Duration

- Intervals

Frequency

Counting # of times behavior occurs

Duration

How long behavior occurs

Intervals

Whether behavior occurred within an interval

Types of Observation Studies

- Naturalistic (unobtrusive)

- Participants

naturalistic observation method

- observing and recording behavior in naturally occurring situations without trying to manipulate and control the situation

- hidden observation

participant observation method

- Interacting with group of interest

Types of non-experimental designs

- survey

- observational research

- archives

- Case study

face validity

Measures whether a test looks like it tests what it is supposed to test.

Test for nominal variable

Chi-square

Test for ordinal variable

Mann-whitney u

Test for interval or ratio variable with two or more IVs

Two-way ANOVA

Test for interval or ratio variable with one IV + 2 levels (within-sub)

Dependent-samples t test

Test for interval or ratio variable with one IV + two level (between-sub)

Independent-samples t test

Test for interval or ratio variable with one IV + three or more levels

One-way ANOVA

double-barreled questions

asking two questions in one

forced-choice questions

A question that asks respondents to choose an answer from possibilities given on a questionnaire.

semantic differential format

a response scale whose numbers are anchored with adjectives

fence sitting

playing it safe by answering in the middle of the scale for every question in a survey or interview

faking good

socially desirable responding

faking bad

giving answers on a survey that make one look worse than one really is

observer bias

observers' expectations influence their interpretation of the participants' behaviors or the outcome of the study

observer effects (expectancy effects)

a change in behavior of study participants in the direction of an observer's expectation

masked design (blind design)

observers are unaware of the conditions to which participants have been assigned and are unaware of what the study is about

unobtrusive observation

an observation in a study made indirectly, through physical traces of behavior, or made by someone who is hidden or is posing as a bystander

Reactivity

A change in behavior of study participants because they are aware they are being watched.

acquiesce

agreeing to every item instead of thinking carefully about it

content validity

the extent to which a measure captures all parts of a defined construct

cross-lag correlations

in a longitudinal design, a correlation between an earlier measure of one variable and a later measure of another variable

temporal precedence

one of three criteria for establishing a causal claim, stating that the proposed causal variable comes first in time, before the proposed outcome variable

concurrent-measures design

participants are exposed to all the levels of an independent variable at roughly the same time, and a single attitudinal or behavioral preference is the dependent variable

demand characteristics

a cue that leads participants to guesas a study's hypothgeses or goals

large n + small amount of unsystematic variability =

power

stable-baseline design

A small-N design in which a researcher observes behavior for an extended baseline period before beginning a treatment or other intervention; if behavior during the baseline is stable, the researcher is more certain of the treatment's effectiveness.

multiple baseline design

a small-N design in which researchers stagger their introduction of an intervention across a variety of contexts, times, or situations

reversal design

A small-N design in which a researcher observes a problem behavior both before and during treatment, and then discontinues the treatment for a while to see if the problem behavior returns.

principle of beneficence

researchers must take precautions to protect research participants from harm and to ensure their well-being

Cronbach's alpha

a correlation-based statistic that measures a scale's internal reliability

convergent validity

an empirical test of the extent to which a measure is associated with other measures of a theoretically similar construct

convergent validity

an empirical test of the extent to which a measure is associated with other measures of a theoretically similar construct

discriminant validity

an empirical test of the extent to which a measure does not associate strongly with measures of other, theoretically different constructs

criterion validity

an empirical form of measurement validity that establishes the extent to which a measure is correlated with a behavior or concrete outcome that it should be related to

purposive sampling

a biased sampling technique in which only certain kinds of people are included in a sample