Slides 5

Question 1

what is the gold standard in psychology experimental designs

Answer 1

random selection

• every individual I that population has the same chance to be selected to participate in that research study
• improves external validity

Question 2

why isn’t true random selection performed often if its the gold standard?

Answer 2

rarely feasible unless the population is very narrowly defined

Question 3

when using convenience samples, what should the investigator do?

Answer 3

the investigator should evaluate and at least discuss whether unique features of the sample contributed to the results

Question 4

what is equivalence of groups in a study dependant on? why?

Answer 4

N

• even if participants are randomly assigned groups, if your sample is very small, say 10, it could be that by chance one group is different than the other
• also: in small groups, there isn’t enough power there to determine if your groups are significantly different, so really, you’re in the dark

Question 5

what to do to make sure groups are equivalent if sample of participants for study is small

Answer 5

match participants

• guarantees equivalence on key variables
• of course have to be cautious with this bc sometimes you may think you’re matching participants when in fact you aren’t
  ex: multiple ways to get same IQ score

Question 6

If the midterm includes an experimental design that says its nonrandom, what is it likely to be ?

Answer 6

quasi-experimental design

Question 7

what does R mean in experimental design notation

Answer 7

random assignment

Question 8

what does O mean in experimental design notation

Answer 8

observation/assessment

Question 9

what does X mean in experimental design notation

Answer 9

treatment or manioulation

Question 10

what is this design called

R O1 X O2
R O3 O4

What are it’s strengths? Weaknesses?

Answer 10

Pretest-posttest control group design

Strengths:
is period between pretest and posttest are equivalent than many threats to interval validity are controlled for

• allows for matching using pretest scores
• increased power, better examination across groups
• allows for predictors and descriptors of attrition

Weaknesses:
pretest sensitization
- did exposure to the pretest contribute to the response of the manipulation?

Question 11

what is this design called

R X O1
R O2

What are it’s strengths? Weaknesses?

Answer 11

posttest only control group design

strengths:
pretest sensitization
- did exposure to the pretest contribute to the response of the manipulation?

weaknesses
is period between pretest and posttest are equivalent than many threats to interval validity are controlled for

• allows for matching using pretest scores
• increased power, better examination across groups
• allows for predictors and descriptors of attrition

Question 12

what is this design called

R O1 X O2
R O3 O4
R X O5
R O6

what does it do?

Answer 12

Soloman Four-group design

evaluates the effect of the pretest

• typically only really used when we are sure that pretesting will produce problems

Question 13

what is a factorial design

Answer 13

allows for simultaneous investigation of the effect of two or more variables

allows for testing of interactions

Question 14

what is this design called

nonR O1 X O2
nonR O3 X O4

what does it do?

Answer 14

Quasi-experimental design

• includes pretest, posttest, and other variations
• doesn’t have random assignment

Question 15

Why might an experiment have nonrandomly assigned subjects

Answer 15

subjects who are already in separate clinics, schools, or classrooms
- there is a reason they cannot be compared

Question 16

what is this design called

R O1 XA O2 XB O3
R O4 XB O5 XA O6

what does it do?

Answer 16

crossover design

• subjects differ only in the order they receive tx
• design is optimal for 2 randomly assigned groups

Question 17

what happens for multiple treatments counterbalanced designs

Answer 17

in crossover deisgns, the order of treatments are counterbalanced

increased number of treatments becomes more difficult

Question 18

what is it called when an experimental design has sequences arranged where each treatment occurs in a positive only once (kind of like suduko)

Answer 18

a latin square

A B C D
B A D C
C D A B
D C B A

Question 19

what is not good about latin quare designs

Answer 19

not every treatment is proceeded and followed by every other treatmet so its possible there is an interaction between a treatments affectiveness and its order to other treatments

Question 20

What happens in a true counterbalanced design

Answer 20

there is random assignment to every possible sequence of treatments

Question 21

what are sequence effects

Answer 21

arrangement of treatments contributes to effects - like carry over effects

Question 22

what does a no treatment control group accomplish

Answer 22

helps rule of confounds

-

Question 23

how do you assign your no treatment control groups

Answer 23

randomly assign potential subjects

- cannot use drop outs as controls (they are self selected as opposed to randomly selected

Question 24

what often happens with no treatment control groups

Answer 24

may seek treatment elsewhere

they often drop out

Question 25

what are waitlist control groups

Answer 25

use subjects who are waiting for treatment as a control group

treatment is given after treatment is over for group 1
- must only use participants who in advance agree to remain if placed in this group, otherwise could introduce serious confounds

waitlist group completes 2 rounds of pre-assessment and one of post

Question 26

what are no contact control groups

Answer 26

they don’t even know they are in the study

• pre-test info from another sutyd or a large screening
• ethical issues
• subsequent test info obtained through a ruse so they don’t know its related to the study

Question 27

what are nonspecific-treatment or “attention-placebo” control groups

Answer 27

nonspecific factors are controlled so the specific mechanisms of change can be somewhat isolated and evaluated

• control for common factors
• many ethical issues
• reduced faith in therapy
• leads individuals to never seek real treatment
• problem with withholding treatment

Question 28

what are routine/standard treatment control groups

Answer 28

all receive an acceptable treatment

limited attrition
controls for nonspecific factors
more likely to have therapist compliance

Question 29

what are yoked control control groups

Answer 29

control subjects are matched with subjects to complete aspects of the protocol
- issues with construct validity should drive this decision (rule out demand characteristics)