PSYC2030 Exam Review #2 March 13 th 2014. Chapter 5.

PSYC2030Exam Review #2

March 13th 2014

Chapter 5

Essential Features of Experimental Research

• What is an independent variable and what is meant by the term “levels” of an independent variable? Independent variable:

The factor of interest to the researcher Must have a minimum of two levels

(Situations/Conditions) Can be manipulated by the experimenter

E.g., experimentally eliciting anxiety in Group A but not in Group B

Participants can be selected based on their attributes E.g., selecting two groups that differ in

anxiety level


• Distinguish between an experimental and a control group. Experimental group:

Receives treatment Control group:

Treatment is withheld = comparison group


• What is a confounding variable and why is it a problem in experimental research? Confounding variable:

Any uncontrolled extraneous variable that covaries with the independent variable and could provide an alternative explanation of the results

Problem?Difficult to interpret resultsResults could be due to the effect of the

independent variable, confounding variable, or some combination of both


• In psychological research, what is a dependent variable? Behaviour measured as an outcome of an

experiment If we study the effect of X on Y

X = Independent Variable Y = Dependent Variable


• What is the difference between a ceiling and a floor effect? Ceiling effect

The average score in both groups is so high that no difference can be determined E.g., questions are too easy

Floor effect The average score in both groups is so low

that no difference can be determined E.g., questions are too difficult

Manipulated versus Subject Variables

• What is meant by a subject variable? Give two examples. Subject variable: Existing characteristics

of the individual participating in the study Examples

Sex Age Socioeconomic status

Manipulated versus Subject Variables

• When drawing conclusions from studies using manipulated variables, statements about cause can be made. These statements cannot be made when using subject variables. Why? With manipulated variables, groups are created based

on random assignment The researcher can be fairly confident that the groups

are are equal (i.e., any differences will be random) With subject variables, groups are created based on

participants’ characteristics The researcher can’t be certain that the groups are equal

(i.e., differences might not be random) Any differences might be due to individual differences

than covary with the subject variables

The Validity of Experimental Research

• Give two examples of mistakes that result in a loss of statistical conclusion validity. What is statistical conclusion validity?

The extent to which the researcher uses statistics properly and draws the appropriate conclusions from the statistical analysis

Examples of mistakes: Using the wrong analysis technique or violate

some of the assumptions required for performing a particular analysis

Reporting some results but not others

The Validity of Experimental Research

Distinguish internal from external validity. Internal validity: the degree to which an

experiment is methodologically sound and confound-free

External validity: the degree to which research findings generalize beyond the specific experiment being conducted

Threats to Internal Validity• How might attrition reduce an experiment’s internal

validity? Attrition: participants leaving the study over the

course of time Problem?

The group finishing the study is made up of different types of people than is the group that started the study.

Threats to Internal Validity• Distinguish between testing and instrumentation

as threats to the internal validity of a study. Testing: Occurs when the fact of taking a pretest

influences posttest scores (perhaps by sensitizing participants to the purpose of the study)

Instrumentation: Occurs when the measuring instrument changes from the pretest to the posttest

Chapter 6

Between-Subjects Designs In a between-subjects design, why is it

important to aim for equivalent groups? Between subjects design: an experimental

design in which different groups of participants serve in the different conditions of the study

we want to make sure that any differences between the groups is due to the levels of the independent variable

Between-Subjects Designs Other than the problem of equivalent groups,

what is a disadvantage of using a between-subjects design? Large numbers of people may be need to be

recruited, tested, and debriefed during the course of the study

The Problem of Creating Equivalent Groups

What is the difference between random selection and random assignment? Random selection: A procedure for getting

volunteers to come into your study. Every person has an equal chance to participating.

Random assignment: A procedure to assign participants to conditions. Every person participating has an equal chance of bring placed in any of the conditions/groups in the study.


Explain how random assignment accomplishes the goal of creating equivalent groups Individual differences are randomly spread

out across conditions


Explain how simple random assignment differs from blocked random assignment With simple random assignment, there is

no guarantee that the groups will be equal in size (e.g., flipping a coin 20 times can result in heads all 20 times)

With blocked random assignment, participants are assigned to blocks of alternating conditions

Within-Subjects Designs

What are the advantages of within-subjects designs over between-subjects designs? Within-subjects designs: an experimental

design in which the same participants serve in each of the conditions of the study

Advantage over between-subjects designs?Fewer people are required!Eliminate the equivalent groups problem


What is the main problem that a within-subjects design has to solve? Order effect: the experience of

participating in one condition of the study influences performance in subsequent conditions


What is the difference between a progressive effect and a carryover effect? Progressive effect: performance changes

steadily from trial to trial (e.g., fatigue) Carryover effect: performance changes as

a result of one sequence of events rather than another (e.g., easy-difficult Qs vs. difficult-easy Qs)

The Problem of Controlling Sequence Effects

• What is the general purpose of counterbalancing? The order of conditions/presentation is

randomized for each subject, thus eliminating order effects


what is “complete” counterbalancing? If there are five different conditions in a study, how many sequences will be needed if complete counterbalancing is going to be used? complete counterbalancing occurs when

all possible orders are used Five conditions: 5! = 120 sequences


• What is “partial” counterbalancing? What are two ways to accomplish partial counterbalancing? Partial counterbalancing: a subset of all

possible orders of conditions is used Sampling from the complete set of all

possible orders Randomizing the order of conditions for

each subject

Control Problems in Developmental Research

What is a cross-sectional design? In developmental psychology, a design in

which age is the independent variable and different groups of people are testedEach group is of different age


What are cohort effects and how might they make the interpretation of cross-sectional designs difficult? Cohort: group of people born at the same

time Cohort effects: differences between

groups are due to growing up in different historical times


What is a longitudinal design? In developmental psychology, a design in

which age is the independent variable and the same group of people is being tested repeatedly at different ages


Describe how a cohort sequential design combines elements of cross-sectional and longitudinal designs A new cohort is added to a study every few

years and then studied periodically throughout the time course of the study

Attempts to balance cohort and attrition problems

Problems with Biasing What is experimenter bias and how can it be

demonstrated experimentally? Occurs when an experimenter’s

expectations about a study affect its outcome

Problems with Biasing What are two ways to control experimenter

bias? Research protocols: detailed descriptions

of of the sequence of events in a research sessionEnsures uniformity of treatment of research

participants A double blind procedure: neither the

participant nor the experimenter knows which condition of study is being tested

Problems with Biasing What is the Hawthorne effect?

A form of participant bias in which a participant’s behaviour is influenced by the mere knowledge of being in an experiment and therefore important to the experimenter

Problems with Biasing• What is the good subject effect, how does it relate

to demand characteristics, and how might it influence the outcome of a study? Good subject effect: A form of participant bias in

which participants try to guess the experimenter’s hypothesis and then behave in such a way to confirm it

Demand characteristics: A feature of the experimental design that increases the chance that participants will detect the true purpose of the study

Problems with Biasing• How can participant bias be controlled?

Reduce demand characteristics Deception Placebo control group

Manipulation check: Asks participants what they thought the

study was about Checks for the presence of demand

characteristics Conduct field research

Chapter 7

Single-Factor – Two Levels• Describe the Stroop effect and explain why his

first two studies were examples of single factor repeated-measures designs. Single factor design: An experimental

design with a single independent variable Stroop effect: When the name of a color is

printed in a color that doesn’t match the name, naming the color of the word takes longer and is more prone to errors than when the color of the ink matches the name of the color

Single-Factor – Two Levels Identify the situations when a t test for

independent samples will be used and when a t test for related samples will be used. Independent t-test

Used when two groups are completely independent from one another Random assignment Subject variables with two groups (males vs. females)

Dependent samples t-test: Matched groups design Repeated-measures designs

Presenting Data There are Three Ways: 1. Numbers can be presented in

sentence form 2. Construct a table of results 3. Graphical Representation

Presenting Data • How you present your data depends on the

type of IV you are interested in... Continuous Data: Use a line graph, but a

bar graph is adequate Discrete Data: Use a bar graph, line graph

is inappropriate. The IV is always presented on the X axis

or the horizontal line The DV is always presented on the Y axis

or vertical line

Control Group Designs• What is a placebo and what is the reason for

including a placebo control group in an experiment? Placebo: a substance given to a

participant in a form suggesting a specific pharmacological effect when in fact the substance in inactive

Purpose? Sometimes effect is caused simply by believing in the drug.

Control Group Designs Placebo Control Groups: Participants

are led to believe they are expecting some treatment when they are actually not. This type of control group is often used in pharmacological studies

Control Group Designs Waitlist Control Groups: Often used

to assess the effectiveness of the program and helps eliminate alternative explanations for the results you obtain such as general improvement because of passage of time

Control Group Designs Yoked Control Groups: when each

member of the control group is matched or “yoked” to a member of the experimental group so that the amount of time spent participating in the events in each condition is equivalent across your groups.

PSYC2030 Exam Review #2 March 13 th 2014. Chapter 5.

Documents