Experiments

Question 1

Experiment

Answer 1

A systematic research study in which one variable manipulates another variable

Question 2

Independent Variable

Answer 2

The manipulated variable that influences the behavior of the dependent variable

Question 3

Dependent Variable

Answer 3

The variable that is being observed; changes are based on manipulating the independent variable
- NOT MANIPULATED

Question 4

What happens when no variables are manipulated?

Answer 4

It is not an experiment
- One variable needs to be manipulated (have 2 levels/conditions) to be considered an experiment

Question 5

What 2 factors are not a requirement for a study to be an experiment?

Answer 5

1. Having a variable with at least 3 levels
2. Showing that variation in a manipulated variable leads to variation in an observed variable

Question 6

Control Variable

Answer 6

A variable that is controlled for the purpose of holding it constant across all levels of the independent variable

Question 7

Covariance

Answer 7

There should be a difference observed in the DV, due to the levels of IV

• if you want to make a causal claim

Question 8

Temporal Precedence

Answer 8

The manipulated variation of the IV needs to be set up before the DV is even measured

Question 9

Internal Validity

Answer 9

Other causes of variation in the study other than the IV need to be ruled out

• remove confounds

Question 10

Confounding Variable

Answer 10

Threats to internal validity; provide alternative explanation of results

• third variable
• changes in parallel with IV

Question 11

How do we need to view all potential confoudns?

Answer 11

Have to treat every confound as messing up the causal claim

• they could have influenced the DV (cannot determine

Question 12

What are the 3 criteria for establishing causality?

Answer 12

1. Covariance
2. Temporal Precedence
3. Interna Validity

Question 13

Do experiments need a control condition?

Answer 13

No, there needs to be a comparison condition, but there does not need to be a controlled condition

Question 14

What are the 3 types of confounds?

Answer 14

1. Design confounds
2. Selection effects
3. Order effects

Question 15

Design Confound

Answer 15

A variable that varies systematically along with the IV

= a flaw in the study’s design that makes it unclear what actually caused the outcome.

Question 16

Selection Effects

Answer 16

Participants in different conditions of the experiment end up being different
(all male versus all female)

• only occurs when participants in conditions changes

Question 17

Order Effects

Answer 17

The order in which conditions are presented to participants

Question 18

If something is the same across all conditions….

Answer 18

it is not a confound

Question 19

Between Subjects Designs

Answer 19

Different participants in different conditions of experiment

Question 20

What is manipulated in between subjects design?

Answer 20

The independent variable between different groups of subjects

Question 21

Within Subjects Design

Answer 21

Same subjects in all conditions of an experiment
- each participant is exposed to every condition of the experiment

Question 22

What is manipulated in within subjects design?

Answer 22

Independent variable within a single group of subjects

Question 23

What confound can only occur in between subjects design?

Answer 23

Selection Effects

Question 24

What confound can only occur in within subjects design?

Answer 24

Order Effects

Question 25

What are the 2 types of between subjects experiments?

Answer 25

1. Independent Groups Post Test Only - most common 2. Independent Groups Pre/Post Test

Question 26

Independent Groups Post Test Only

Answer 26

Manipulate Independent Variable and measure the dependent variable

Question 27

Independent Groups Pre/Post Test

Answer 27

Measure dependent variable before manipulating independent variable, and then measure dependent variable after manipulation

Question 28

How do you control the confound: Selection Effects?

Answer 28

1. Random Assignment 2. Matching

Question 29

Random Assignment

Answer 29

Each participant has an equal chance of being placed in any experimental group (e.g., treatment vs. control).

Question 30

Matching

Answer 30

Participants are paired or grouped based on a key variable (e.g., age, IQ, gender), and then assigned to conditions so each group has similar values on that variable. - make conditions similar in each if it is a possible confound

Question 31

What is a con about matching?

Answer 31

Complex - need to identify more information about participants just to sort them into conditions

Question 32

When can random sampling be conducted?

Answer 32

Large sample size (15; 30 is more ideal)

Question 33

Demand Characteristic

Answer 33

Something about the experiment gives away what the experiment is trying to measure - only do pre-test if no demand characteristics

Question 34

What methods are conducted when the sample size is not large?

Answer 34

1. Pre test/post test - no demand characteristics present 2. Matching - only if logistically feasible and valid

Question 35

What are the 2 types of within subject design?

Answer 35

1. Concurrent Measures 2. Repeated Measures

Question 36

Concurrent Measures

Answer 36

Participants receive both levels of independent variable at the same time - comparing words with 2 dif font types

Question 37

Repeated Measures

Answer 37

Participants receive the different levels of independent variables in different trials - first trial is one font, the next is another font = dependent variables are measured repeatedly over time

Question 38

Counterbalancing

Answer 38

Presenting the levels of the independent variable to participants in different sequences to control for order effects

Question 39

When is counterbalancing used?

Answer 39

Within subjects design when more than one order is being used for the same group of participants

Question 40

What are the 2 ways to counterbalance?

Answer 40

1. Full counter balancing 2. Partial Counter balancing with random ordering

Question 41

Full Counter balancing

Answer 41

All possible orders are used across different subjects - K! possible orders

Question 42

Partial Counter Balancing

Answer 42

Each subject gets the conditions in a different order

Question 43

Construct Validity

Answer 43

How well the variables were measured and manipulated

Question 44

External Validity

Answer 44

To whom and to what settings you can generalize the causal claim

Question 45

Internal Validity

Answer 45

Are there alternative explanations for the outcome recorded - potential confounds?

Question 46

Statistical Validity

Answer 46

How well the data supports the causal conclusion

Question 47

How do you test statistical validity if there are 2 independent variable conditions?

Answer 47

T test for differences between sample means

Question 48

T Test for independent groups

Answer 48

Between subjects design or nonmatched groups

Question 49

T Test for dependent groups

Answer 49

Within subjects design or matched groups

Question 50

What type of claim is most powerful?

Answer 50

CAUSAL CLAIMMMMMMMm