quiz 3 psych206

What are the 3 key features of an experimental design?

independent variable
dependent variables
control variables

independent variable

manipulated variable

dependent variable

what changes as a result of manipulating the IV

control variables

holding another variable "constant" in an experiment means that you make the variable the same across all conditions

Why are you most confident of causality in experimental studies versus descriptive studies or correlational studies?

Because you've got more control in an experiment. You can rule out other reasons for why your dependent variable changed. In other words, you're more confident that the change in the IV caused the change in DV, because you've gotten rid of other potential

What are the three criteria for causality?

covariation, temporal precedence, internal validity

Covariation

one variable changes when another variable changes

temporal precedence

the one variable that you think affects the other variable always occurs earlier in time

internal validity

you can rule out other explanations for why the DV changes

What, specifically, is a design confound and what does it mean for your study?

a. A design confound is a mistake in the study design that allows a second variable to vary systematically along with the intended Independent variable. It means that you've now got an alternative explanation for why the dependent/measured variable change

What is the difference between systematic and unsystematic variability?

a. In systematic variability, another variable varies systematically between the conditions. Systematic variability undermines internal validity - it provides another reason for the DV changing. In unsystematic variability, another variable may influence

What's a selection effect? Why is it a problem?

a. A selection effect occurs when participants in one condition are systematically different than participants in another condition. This is a problem because it creates an alternative explanation for why the DV differs between conditions - it's not becau

What's "random assignment"?

a. In random assignment, researchers randomly put people into the different conditions of their experiment. They do this to avoid selection effects; say you want to do a study on fear and spatial reasoning. You want to make sure that people's baseline ind

1. I show one group of people sad faces very quickly (<.5 sec), with each face followed by their name. I show another group of people happy smiling faces very quickly (<.5 sec), with each face followed by their name. Afterward, I give each group a questio

independent groups

independent groups

different groups of people experienced different levels of the independent variable

I show one group of people sad faces very quickly (<.5 sec), with each face followed by their name. I show another group of people happy smiling faces very quickly (<.5 sec), with each face followed by their name. Afterward, I give each group a questionna

type of face

I show one group of people sad faces very quickly (<.5 sec), with each face followed by their name. I show another group of people happy smiling faces very quickly (<.5 sec), with each face followed by their name. Afterward, I give each group a questionna

self-esteem

What should I do to make the above study a pre-test/posttest design?

Add a questionnaire that measures self-esteem before and after showing the faces.

In thinking about the construct validity of a study, what's a question you want to ask yourself about the dependent variable

how did they measure it

In thinking about the construct validity of a study, what's a question you want to ask yourself about the independent variable?

How well did the manipulation work? (is the manipulation face valid [does it make sense?]; have others used it?)

What's external validity? what's a question you want to ask about the sample?

How well does it represent the population you want to generalize too?

Also in thinking about external validity, what's a question you want to ask about the situation?

How well does it generalize to other situations?

Let's say the self-esteem scores of the groups from question 7 are pretty different from each other. Thinking about statistical validity, what are two questions you want to keep in mind?

a. Was the difference significant
b. What was the effect size

Max ran an experiment in which he asked people to shake hands with an experimenter (played by a female friend) and rate the experimenter's friendliness using a self-report measure. The experimenter was always the same person, and used the same standard gr

Because the participants shook the experimenter's hand before rating her friendliness

Max ran an experiment in which he asked people to shake hands with an experimenter (played by a female friend) and rate the experimenter's friendliness using a self-report measure. The experimenter was always the same person, and used the same standard gr

the standard greeting the experimenter used with everyone when shaking hands.

Max ran an experiment in which he asked people to shake hands with an experimenter (played by a female friend) and rate the experimenter's friendliness using a self-report measure. The experimenter was always the same person, and used the same standard gr

Because he wanted to avoid selection effects.