Quiz 3

Question 1

true experiment

Answer 1

research procedure in which the scientist has complete control over all aspects

Question 2

quasi experiment

Answer 2

research procedure which does not meet the requirements of a true experiment

Question 3

factor

Answer 3

the independent variables of an experiment

Question 4

level

Answer 4

in an experiment, a particular value of an independent variable

Question 5

Condition

Answer 5

a group or treatment in an experiment

Question 6

Treatment

Answer 6

another word for a condition of an experiment

Question 7

between

Answer 7

subjects design

Question 8

within

Answer 8

subjects design

Question 9

order effect

Answer 9

changes in a subject’s performance resulting from the position in which a condition appears in an experiment

Question 10

sequence effect

Answer 10

changes in a subject’s performance resulting from interactions among the conditions themselves

Question 11

block randomization

Answer 11

control procedure in which the order of conditions is randomized but with each condition being presented once before any condition is repeated

Question 12

Counterbalancing

Answer 12

controlling for order and sequence effects by arranging that subjects experience the various conditions in different order

Question 13

Reverse counter balancing

Answer 13

method of control in which conditions are presented in order the first time and then in reverse order

Question 14

Latin square

Answer 14

control procedure in which each subject experiences each condition in a different order from other subject

Question 15

two-conditions design

Answer 15

the simplest research design, involving only two conditions

Question 16

multiple-conditions design

Answer 16

research design that involves more than two conditions

Question 17

one-group posttest-only design

Answer 17

research design that measures the behavior of a single group of subjects after they are given a treatment

Question 18

nonequivalent control group

Answer 18

a group of subjects that is not randomly selected from the same population as the experimental group

Question 19

Define true experiment.

Answer 19

An experiment is a true experiment is when there are a great number of threats that are controlled. It is also a true experiment when the experimenter believes they have control over the instructions that the participants need to carry out; as well as how those instructions are carried out to the participants. When it does not meet these expectations it is known as a quasi experiment. A true experiment has complete control over its variables.

Question 20

Define and distinguish the terms factor, level, and condition when referring to independent variables.

Answer 20

Factors are also known as the independent variable of the experiment or the variables that are being manipulated. The level is a certain value of the independent variables. There are at least two levels to an independent variable (ie. it can experience a certain condition, or it doesn’t experience the condition). The condition is the certain way in which the participants are treated, it is also called a treatment.

Question 21

Describe the two basic elements of good experimental design.

Answer 21

The two basic elements to a good design are having a control group or condition and the randoms assignment of participant to the groups. Except when its a within-subjects-design.

Question 22

What are within-subject designs? What is the basic strategy for achieving control?

Answer 22

In a within-subjects design each participant would act as the control and treatment group. At the least these types of experiment design should randomize the order or counterbalance the order in which the participants experience the condition to achieve more control

Question 23

What are order effects and sequence effects? How are they different?

Answer 23

When a within-subjects design is being used then there is chance that there will be other variables to take into consideration that occur due to the different testing conditions. These effects are called carryover effects. There are two types of carry-over effects. There are order and sequence effects. Meaning, the performance of a participant will be affected by the order in which they experience the conditions for the order effect. Sequence effects occur due to the interactions between the different conditions. The difference between order and sequence effects occurs when there are a number of different ways to order the conditions. Order effects can be controlled by making sure that each condition occurs equally as often for each position it can be in. This is called counterbalancing. Sequencing differs in control in that each condition follows every other condition equally as often

Question 24

How do you control for order and sequence effects within subjects?

Answer 24

To control for order and sequence effect in the within subjects design randomization can be used when the condition is exposed to the participants many times, or when the size of the sample is large enough to not make a big enough difference. Block randomization can also be used to control the sequence and order effect. This is when each participant is randomly picked to create and order, as long as each participant has experienced the condition once before repeating it.

Question 25

When is reverse counterbalancing ineffective in controlling for order effect?

Answer 25

Reverse counter balancing is useful when you don’t have many participants to test, but there are many different conditions they need to be exposed to only a couple times. Reverse counterbalancing presents the condition in order the first time, then presented in reverse order.

Question 26

How do you control for order and sequence effects within groups?

Answer 26

Within group accounts for order and sequencing effects by using complete counterbalancing

Question 27

What is complete counterbalancing? What does it control for? What is its greatest disadvantage?

Answer 27

Complete counterbalancing is when each condition occurs the same number of times for each position and also follows every other condition an equal amount of times. This controls the order and sequence within the group of participants. The disadvantage is that when the amount of conditions gets higher the number of different order arrangements also increases drastically.

Question 28

What is the Latin Square technique? When is this design preferred over complete counterbalancing?

Answer 28

Latin square technique is a tactic used that involves each participant receiving the conditions in a different order than the other participants and not worrying about If each condition follows another condition the same amount of times. This is preferred because less participants are require as opposed to the complete counterbalancing technique.

Question 29

What are between-subject designs? Explain the basic strategy for achieving control.

Answer 29

When subjects cannot be used as their own control then there will be different groups that receive two or more different treatments. The basic strategy for achieving control is establishing a baseline to go by and determining the treatments from there.

Question 30

When are between-subjects designs preferred over within-subjects designs?

Answer 30

Between subject designs are preferred if the within subjects design would not be effective and valid if the participants acted as their own control group.

Question 31

Describe three designs to avoid in research, and give reasons for avoiding them.

Answer 31

One design to avoid is the one group post-test design. This involves monitoring the behaviour of those participant who’s just went through a treatment. This type of design is not useful in any way due to the extra lack of validity from having no control, and establishing no baseline to go by. The second type of design to avoid is the post-test only design with nonequivalent control groups. This is when the participants are not randomly selected from the same population as the experimental group. This type of design is no good because the control group is not representative of the experimental group so this will give invalid results. The other design to avoid is one-group pretest post-test design. This is when a group of participants behaviour is monitored before and after the treatment. This is unfavourable because it’s still doesn’t have the control over the variables as a proper experiment should. There could be many reasons for the changes, or no changes in behaviour that occurs after the treatment.

Question 32

factorial design

Answer 32

research design that involves all combinations of at least two values of two or more independent variables

Question 33

main effect

Answer 33

in a factorial experiment, the effect of one independent variable, averaged over all levels of another independent variable

Question 34

interaction effect

Answer 34

when the effect of one independent variable depends on the level of another independent variable

Question 35

antagonistic interaction

Answer 35

interaction in which the two independent variables tend to reverse each other’s effects

Question 36

synergistic interaction

Answer 36

interaction in which the two independent variables reinforce each other’s effects

Question 37

ceiling-effect interaction

Answer 37

interaction in which one variable has a smaller effect when paired with higher levels of a second variable

Question 38

mixed factorial designs

Answer 38

designs containing at least one variable that is presented in a within-subjects fashion and one other variable that is presented in a between-subjects fashion

Question 39

What are factorial designs?

Answer 39

Factorial designs are when all combinations between two or more values of two or more factors are used.

Question 40

What are the advantages of using factorial designs?

Answer 40

This is useful when one is trying to determine if a combination of variables rather than a single variable are responsible for the changes of behaviour or whatever whats being measured.

Question 41

How many cells (combinations) are there in a 2 x 3 x 4 factorial design?

Answer 41

24

Question 42

What are main effects?

Answer 42

The main effect is the amount of influence one factor has on the dependent variable. If there two or more variable then there is a possibility of two or more main effects.

Question 43

What are interaction effects? How can one tell graphically if there is an interaction between two variables?

Answer 43

Interaction effects are the amount of influence one factor has on another. Meaning the influence of one factor depends on how much the other factor is being used. One can tell if there is an interaction between two variables if the curves of the graph lines are not parallel.

Question 44

What are the common types of interactions in psychological research?

Answer 44

The common types of interactions in psychological research are antagonistic, synergistic, and ceiling effect. The antagonist interaction does not have any main effects. The effects of the independent variables tend to reverse each others. Synergistic interaction is when the factors help along each others influence. The ceiling effect occurs when one of the variables has less influence when it is coupled with more of the second variable.

Question 45

What is the difference among within-subjects, between-subjects, and mixed factorial designs?

Answer 45

Within-subjects design is when all participants receive all the same multiple treatments, therefore, all participants are tested in all possible conditions. When subjects can’t act as their own controls in the factorial design then the between-subjects design is used. This is when separate groups with a varying amount of subjects will each be given a different combination of treatments until all the combinations are used. The mixed factorial design is when both the within subject and between subject are both used for different variables.

Question 46

quasi experiments

Answer 46

research procedure in which the scientist must select subjects for different conditions from pre- existing groups

Question 47

non-equivalent control group designs

Answer 47

research design having both an experimental and a control group wherein subjects are not randomly assigned to groups

Question 48

repeated treatment designs

Answer 48

a design in which a treatment is withdrawn and then presented a second time

Question 49

interrupted time-series designs

Answer 49

research design that allows the same group to be compared over time by considering the trend of the data before and after experimental manipulation

Question 50

cohort

Answer 50

in research, a group that has something in common, such as age

Question 51

secular trend

Answer 51

a change that is taking place in the general population over time (the term has nothing to do with religion)

Question 52

program evaluation

Answer 52

a set of techniques for determining the effectiveness of a social service program.

Question 53

meta-analysis

Answer 53

a set of methods for combining the results of many studies

Question 54

time-lag effect

Answer 54

in a cross-sequential design, the effect resulting from comparing subjects of the same age at different times

Question 55

summative evaluation

Answer 55

an evaluation of the quality of a project, often after it is completed

Question 56

formative evaluation

Answer 56

an evaluation of ways to improve a project while it is ongoing

Question 57

cross-sectional study

Answer 57

in developmental research, a study that tests different age groups at the same time

Question 58

longitudinal study

Answer 58

in developmental research, a study that tests individuals in a single cohort over the course of time

Question 59

cross-sequential design

Answer 59

design used to help separate developmental, cohort, and secular effects;
tests individuals from two or more cohorts at two or more different times ( a blending of cross-sectional and longitudinal) in order to separate developmental, cohort, and secular effects.

Question 60

How are quasi experiments different from true experiments?

Answer 60

a true experiment is one in which the experimenter has complete control over the who, what, when, where, and how of the experiment. A quasi experiment, by contrast, does not permit the experimenter to control the assignment of subjects to conditions.

in true experiments we manipulate variables, whereas in quasi experiments we observe categories of subjects.

in quasi, we have the additional problem of not knowing whether the difference in behavior was caused by differences between the groups that existed before the experiment started or by the independent variable.

a true experiment permits the most powerful control for confounding because all other potential independent variables have been eliminated by randomly assigning subjects to conditions. A quasi experiment leaves open the possibility that other differences exist between the experimental and control conditions and thus permits other potential differences to remain.

Question 61

Under what circumstances can quasi experiments be a useful tool?

Answer 61

One good reason for conducting quasi-experiments is that it may be the only possible way to study the impact of a natural event, for example, how residents cope after a flood. For ethical and practical reasons, the researcher cannot manipulate the occurrence of such an event and cannot randomly assign residents to flood versus no flood groups. In cases where the impact of a treatment program or intervention is introduced, the researcher is seldom able to assign clients randomly to treatment versus control groups. The research data collected in quasi-experiments can serve as a rich source of testable hypotheses to be tested in a carefully controlled laboratory experiment. Conversely, the external validity of laboratory findings can sometimes be assessed using the quasi-experimental approach.

Question 62

give an example of a quasi experiment

Answer 62

you may want to study the effect of number of food pellets on the rate at which rats learn a maze. This situation would permit the design of a true experiment because you could arbitrarily assign some rats to the large-reward condition and others to the small-reward condition. Assume that before the experiment the rats belonged to a homogeneous population of rats. For experimental purposes, you assign the rats to groups that you create according to your needs.
However, suppose you suspected that there were sex differences such that the maze running of male rats might be affected differently than female rats by the number of food pellets. In that case, you would have to conduct a quasi experiment because you cannot assign participants to the two conditions, male and female; the rats already belong to those groups naturally. In this situation, the researcher cannot create groups of males and females, but instead selects members from preexisting groups.

Question 63

Quasi experiments are sometimes called

Answer 63

ex post facto, or after the fact, experiments because the experiment is conducted after the groups have been formed.

Question 64

common threats to internal validity include

Answer 64

history, maturation, testing, instrumentation, regression to the mean, selection, and mortality.
diffusion or imitation of treatment occurs when two intact groups share information. Second, compensatory equalization occurs when service providers provide access of treatment services to the control group. The effect of these actions is to reduce the difference between the groups, and the outcome is that the treatment or intervention may be seen as less effective than it should be. Third, compensatory rivalry occurs when the disadvantaged group is motivated to perform better during the experimental period. On the contrary, resentful demoralization can occur when the disadvantaged group performs worse during the experimental period.

Question 65

What is the major weakness of quasi experiments?

Answer 65

The lack of control, either in applying the independent variable or randomly assigning subjects to control & experimental groups

Question 66

you should try first to design a true experimental study and use the other designs only when

Answer 66

you believe that the gain in validity will be worth the loss of control.

Question 67

the best research method is one

Answer 67

that answers the research question

Question 68

What is a non-equivalent control group design with pretest and posttest?
Under what circumstances are the results of quasi-experiments that use a non-equivalent control group design interpretable?

Answer 68

A non-equivalent control group is a group of subjets that is not randomly selected from the same population as the experimental group. any behavioral change from pretest to posttest of an experimental group is compared with that of a non-equivalent control group. Two result patterns of this design are generally interpretable: (1) if the behaviour of both groups are the same on the pretest and the experimental group’s behaviour has changed on the posttest while the control group’s has not, or (2) if there is a crossover between the experimental group and the control group on the pretest and on the posttest (i.e., the experimental group has a lower measure than the control group on the pretest but a higher measure on the posttest).

Question 69

The problem with a non-equivalent control group is

Answer 69

determining how to compare results between the experimental and control groups when the two groups were not equivalent to begin with

Question 70

Describe a study using a mixed factorial design with one non manipulated variable

Answer 70

“Edmund Keogh and Gerke Witt (2001) thought that caffeine intake might influence the perception of pain, and that the effect might be different in men than in women. They tested this theory by having 25 men and 25 women take part in two sessions separated by a week. In one of the sessions, participants drank a cup of coffee that contained caffeine and in the other session the coffee was decaffeinated. In each session, participants put their nondominant hand in an ice-water bath and were instructed to indicate the point of just noticeable pain. The design of this experiment is a mixed factorial because it has one between- subjects variable and one within-subjects variable. The between-subjects variable is sex because participants were either male or female and this is a quasi- experimental variable. The within-subjects variable is caffeine because participants experienced both conditions at different times. Caffeine generally produced a higher pain thresh- old, and that women both had a slightly higher pain threshold than men and showed the greatest increase related to caffeine.

Question 71

What is an interrupted time-series design? Under what circumstances is this design useful?

Answer 71

A research design that allows the same group to be compared over time by considering the trend of the data before and after experimental manipulation.

Sometimes no control group can be obtained that can be considered comparable enough to be useful. Then a design that allows the same group to be compared over time can be used. It is useful when no control group is available and the researcher wants to not only look at the difference in a behavioral change but also at trends or changes that take place during significant event.

The interrupted time-series design comprises three phases: (1) a series of pretest measurements over time, (2) the occurrence of the significant event (treatment or intervention), and (3) a series of posttest measurements. Many different patterns of results can emerge from interrupted time-series studies. In general, any change in either the slope or the interval at the time of treatment or intervention would suggest the intervention to be the cause of behaviour change. Problems with data interpretation, however, arise frequently when: (1) the treatment or intervention is gradual rather than abrupt, (2) there is a delay impact of the intervention, (3) the time series is short, and (4) there are incomplete or inaccurate archival data

Question 72

What different strategies are used in non-equivalent control group designs and interrupted time-series designs to control for confounding variables?

Answer 72

Ruling out alternative hypotheses, i.e., comparing similar groups outside of the ‘study’ groups. Taking multiple baseline readings or having access to baseline material from before the independent variable was introduced to rule out maturation, practice, etc.

Question 73

What are repeated-treatment designs? How do these designs control for confounding variables?

Answer 73

A design in which a treatment is withdrawn and then presented a second time. The subject’s response is measured before and after the introduction of a treatment, then the treatment is withdrawn and the whole process is begun again.

Question 74

What is a limitation to the repeated treatment design?

Answer 74

treatment must be one that can be withdrawn without causing complications in the analysis of data.

Question 75

Which quasi-experimental designs can be used with a single subject (an individual, a company, or some other units)?

Answer 75

Repeated treatment & interrupted time-series design can be used with single subjects

Question 76

What are the common quasi-experimental designs used in developmental psychology? What are their relative advantages and disadvantages?

Answer 76

Cross-sectional studies test different age groups at the same time. The disadvantage is that people at different ages were all born at different times, so age is confounded with date of birth. A cohort is a group that has some characteristic in common, and so people with this characteristic are treated as a group. To reduce cohort effects, a longitudinal study, which follows one or several cohorts separately, could be used. However, longitudinal studies take years to complete as the cohort ages (practical problem). The main theoretical problem with a longitudinal study Is that it confounds age with time of testing. Secular trends, or changes taking place in the general population over time, can cause confounds, i.e., technology could be changing, causing differences in performance not related to birth year. A cross-sequential design tests individuals from two or more cohorts at twto or more different times ( a blending of cross-sectional and longitudinal) in order to separate developmental, cohort, and secular effects.

Question 77

What is programme evaluation? How can quasi-experimental designs be used in evaluations of social programmes?

Answer 77

Program evaluation is a set of techniques for determining the effectiveness of a social service program. Evaluators would compare the program with other programs (standard treatments) or outcomes between those who are and are not participating in a program. Program evaluation is of two major types: summative and formative. The goal of summative evaluation is to assess whether a given program is functioning adequately. Its outcome is used to determine whether a program should be cut back, discontinued, maintained, or expanded with additional funding. The objective of formative evaluation is to recommend changes that will improve program functioning. Potential political implications of the research outcome and the impact on a program’s stakeholders may cause resistance to evaluation, especially summative evaluation, to be frequent and fierce. One way to minimize resistance is for the evaluator to establish a rapport with the stakeholders early on in the process. When conducting program evaluations, researchers are well-advised to be cognizant of the politics involved every step of the way. Quasi experimental designs are often used in program evaluations, because clients of the program are more likely to be assigned, rather than to be randomly allocated, to groups for comparison. So, clients might be matched with sometone who has note been served by the program on some key variables, and then the clients and non-clients could be compared to see how well the program is working by using some variant of a non-equivalent control group design.

Question 78

Sources of resistance to program evaluations

Answer 78

fear the program will be terminated
fear of losing control over the program
fear that information will be abused
fear that wrong measures will be used
belief that evaluation is pointless
hopes are too high

Question 79

steps in planning an evaluation

Answer 79

identify the stakeholders
arrange preliminary meetings
decide whether an evaluation should be done
examine literature
determine the methodology
present a written proposal

Question 80

Give a textbook example of a successful program evaluation

Answer 80

William Fisher reports the results of a campaign to reduce pregnancies among undergraduate students at a large public university in Canada. For the five years before the start of the campaign, positive tests for pregnancy made by the Student Health Service were stable at a rate of about 10 per 1,000 female students. The university initiated a series of pregnancy-prevention lectures in the dormitories in 1983. The next year, a specially developed video- tape and booklet on prevention of pregnancy were added to the program. The pregnancy rate dropped to about 7 per 1,000 in the first year of the pro- gram and then dropped further the next year to 6.5, after which it remained at about the same low level for the three later years reported by Fisher. As a control for a possible secular trend in unwanted pregnancies, Fisher notes that the abortion rate for Canada remained steady during the period. In addition, Fisher believes that the pregnancy rate measured by the Student Health Service was a reliable measure of the pregnancy rate because he sees no good reason to expect that the program would cause pregnant women to avoid the health service. The program was purposely designed to be nonjudgmental and to increase the likelihood that pregnant women would go to the Student Health Service. It is likely that the effect of the program was, if anything, larger than the measure indicates.
One reason that this evaluation succeeded where the previous example failed was that an outcome measure was already routinely available to the Student Health Service. Although not all women who suspected they were pregnant would go to the health service for a test, enough of them would go to give a good indication of the pregnancy rate.