When researchers manipulate a variable, they need to make sure they are only varying one thing at a time. Researchers control potential third variables in their studies by holding all other factors constant. Any variable that the experimenter holds constant on purpose is called the control variable.

Control variables are important for establishing internal validity

DV: measured variable

CV: variable held constant to rule out third variables and establish internal validity.

Experiments establish temporal precedence. The ability to establish temporal precedence is a feature that makes experiments superior to correlational designs.

Well-designed experiments establish internal validity. To be internally valid, a study must ensure that the causal variable and not other factors, is responsible for the change in the effect variable.

For any given research question, there can be several possible alternative explanations.

Can be a threat to internal validity

Unsystematic variability can lead to problems; it can obscure, or make it difficult to detect differences in the dependent variable.

Can occur when experimenters let participants chose which group they want to be in. EX: letting parents chose which group they would like their autistic children to be in, the intensive treatment group, or the usual treatment group.

A way of desystematizing the types of participants who end up in each level of the independent variable.

Used to avoid selection effects

This method is particularly useful if the experiment’s design has outliers (exceptionally clever or inept participants) and if the sample groups are small in size.

This method also ensures that the groups are equal on some important variable, such as IQ, before the manipulation of the independent variable.

An experimental design in which different groups of participants are exposed to different levels of the independent variable, such that each participant experiences only one level of the independent variable

AKA between-subjects design or between-groups design

2 Basic forms of independent-groups designs are the posttest-only design and the pretest/posttest design.

An experiment using an independent-groups design in which participants are tested on the dependent variable only once. Also called equivalent groups, posttest-only design.

Researchers might do this if they want to evaluate whether random assignment made the groups equal

This design also works well to track how participants in the experimental groups have changed over time in response to some manipulation.

EX: babies shown male and female faces at the same time. The babies chose to look at the female face more often than not.

EX: participant –> interact with own toddler –> measure oxytocin levels –> interact with new toddler —> measure oxytocin level.

Generally requires fewer participants overall.

Happen when exposure to one level of the independent variable influences responses to the next level of the independent variable.`

Ex: A – B- C, A – C- B, B – A – C, B – C- A, C – A – B, C – B – A

2 conditions = 2 orders

3 conditions = 6 orders

4 conditions = 24 orders

The first row is set up according to a formula, and then the conditions simply go in numerical order down each column

Might not be possible/practical. (such as teaching children to ride a bike. You teach them with method A, you cannot go back to the baseline and teach them with method B.

People might change the way they act if they see all levels of the independent variable. (Demand characteristics)

repeated measures

assurance of equivalence of the groups

requirement of fewer participants

more statistical power

Disadvantages:

Participants cannot be returned to their original state after each condition.

Demand characteristics may result from hypothesis guessing with repeated testing.

Order effects can threaten internal validity.

.50 = medium, or moderate comparable to an r of .30

.80 = large or strong comparable to an r of .50

posttest only

pretest/posttest

concurrent measures

repeated measures

treatment group

placebo group

experimental group

manipulation group

dependent variable

independent variable

control variable

confound variable

treatment group

manipulation group

experimental group

control group

All participants in the 2-minute condition are tested at 8:00 a.m.; those in the 5-minute condition are tested at noon; and those in the 10-minute condition are tested at 4:00 p.m.

The same list is used for each condition and is randomized for each participant.

The three groups are run simultaneously in three different rooms, and the room for each condition is randomly chosen before each group arrives.

Three different experimenters are in charge of administering the task, and they rotate which condition they are administering.

demand characteristic

selection effect

experimenter bias

carryover effect

independent groups

concurrent measures

matched-groups

within-groups

within-groups

repeated measures

pretest/posttest

posttest only

elimination of practice effects

assurance of equivalence of the groups

requirement of fewer participants

more statistical power

order effects

selection effects

practice effects

carryover effects

four

eight

sixteen

twenty-four

matched groups

random sampling

counterbalancing

random assignment

Extraneous differences are held constant across conditions.

Participants cannot be returned to their original state after each condition.

Demand characteristics may result from hypothesis guessing with repeated testing.

Order effects can threaten internal validity.

posttest only

pretest/posttest

concurrent measures

repeated measures

not existent

weak

moderate

strong

depth perception

number of trials

disparity from perfect alignment

eye/eyes used

distance out of alignment of the rods

how long the participant takes to complete the experiment

improvement in performance by the participant

eye/eyes used

measure more than two variables.

measure two variables.

manipulate two variables.

manipulate one variable and measure another.

weak

moderate

strong

This cannot be determined without knowing the number of participants.

independent groups

within-groups

concurrently

as a participant variable

role of the participant

price that the buyer will pay

price that the seller will accept

type of mug

role of the participant

value selected for the mug

type of mug

the range of prices that the buyer and seller each consider

by using a control group

by using matched-groups design

by using random assignment of participants

by using a pretest/posttest design

A pilot study

A Latin square

Counterbalancing

A manipulation check

A pilot study

A Latin square

Counterbalancing

A manipulation check

-The value of the mug is the measured variable

-Selection effects only apply to independent-groups designs

-Theo assigns the participants to their roles, so that is the manipulated variable

-The group that represents “no treatment” or the neutral condition is the control group

-Tetiana is measuring how far out of alignment the rods are, making that the dependent variable.

-Eric could avoid an order effect by having participants consume and rate the drinks in varying orders

-Manipulation checks help researchers measure how well their manipulation worked

-her findings give a d well above 0.80, so the effect size is strong

-For full counterbalancing, all the possible condition orders must be used, which is 24 for four conditions

-The participants are randomly assigned to the groups and only a posttest is administered

Extraneous differences are held constant across conditions.

Participants cannot be returned to their original state after each condition.

Demand characteristics may result from hypothesis guessing with repeated testing.

Order effects can threaten internal validity.

fatigue

selection

practice

boredom

role of the participant

price that the buyer will pay

price that the seller will accept

name of the participant

within-groups

repeated measures

pretest/posttest

posttest only

independent groups

within-groups

concurrently

as a participant variable

All participants in the 2-minute condition are tested at 8:00a.m., those in the 5-minute condition are tested at noon, and those in the 10-minute condition are tested at 4:00 p.m.

The same list is used for each condition and is randomized for each participant.

The three groups are run simultaneously in three different rooms, and the room for each condition is randomly chosen before each group arrives.

Three different experimenters administer the task and rotate which condition they are administering.

pilot studies

checking for statistical significance

manipulation checks

adding additional conditions

depth perception

number of trials

disparity from perfect alignment

eye/eyes used

construct

internal

statistical

external

elimination of practice effects

assurance of equivalence of the groups

requirement of fewer participants

more statistical power

by using a control group

by using matched-groups design

by random assignment of participants

by using a pretest/posttest design

that each participant will have a unique order of the conditions

that the order of the conditions will be randomized for each group

that each condition appears in each position within the order at least once

that all possible orders of conditions will be used

six

ten

twenty

thirty

demand characteristic

selection effect

experimenter bias

carryover effect

dependent

independent

control

confound

not existent

weak

moderate

strong

treatment

comparison

experimental

control

the percentage correct

the number of objects

the length of exposure

the number of trials

distance out of alignment of the rods

how long the participant takes to complete the experiment

improvement in performance by the participant

eye/eyes used

role of the participant

value selected for the mug

type of mug

name of the participant

the percentage correct

the number of objects

the length of exposure

the number of trials

weak

moderate

strong

This cannot be determined without knowing the number of participants.

four

eight

sixteen

twenty-four

construct

internal

statistical

external

independent groups

concurrent measures

matched-groups

within-groups