Research Design And Statistics

Front 1

subjects become tired

Back 1

maturation

Front 2

drug designed to improve fxn adminstrd to severely depressed

Back 2

statistical regression

Front 3

film designed to increase racial awareness shown day after civil rights leader speach

Back 3

history

Front 4

new diet, Ss weighed before/after; scale breaks b4 2nd weigh in

Back 4

instrumentation

Front 5

study to test workshop's effectiveness; Ss took exam previously

Back 5

testing

Front 6

1st 20 Ss assigned to exp. group; 2nd 20 Ss assigned to control grp

Back 6

selection

Front 7

Cues in the exp setting that allow Ss to guess the research hyp.

Back 7

demand characteristics

Front 8

effect that experimenter's expectancy has on results of study

Back 8

Rosenthal effect

Front 9

procedure designed to ensure all Ss have = chance of being chosen to participate

Back 9

random sampling

Front 10

procedure designed to ensure that Ss have = probability of ending up ineach of the tx groups

Back 10

random assignment

Front 11

tendency of Ss bx to change due to attention received

Back 11

Hawthorne effect

Front 12

stat. method of controlling for the effects of extraneous variable

Back 12

ANCOVA

Front 13

procedure that involves grouping Ss similar on extraneous var. and assigning members of each grp to diff. tx grps.

Back 13

matching

Front 14

population is divided into sub-pops; all members of each sub-pop have = probability of being chosen to participate

Back 14

stratified random sampling

Front 15

Major threat to internal validity of a one-group time-series design:
a. maturation
b. regression to the mean
c. history
d. testing

Back 15

C. one-grp, time-series design involves admin. multiple pretests and post-tests to 1 grp of Ss before and after tx admin. Design controls for many threates to int. val., such as maturation, tesing and stat regression.

Front 16

Greatest threat to validity when a mail survey is conducted?
a. selection
b. maturation
c. randomization
d. instrumentation

Back 16

a. Ss self-select, they decide who study participants will be.

Front 17

Which statements regarding developmental research (longitudinal, cross-sect, cross-seq.) is most true?
a. Cross-seq and Long. studies are particularly vulnerable to "cohort" effects.
b. Cross-seq studies are more costly to conduct (time, money) than long. studies
c. (of the 3 types) cross-sec studies offer the greatest internal validity
d. combining methodology of cross-sec & long, cross-seq reduce many of the problems associated with both

Back 17

d. cross-seq, like cross-sec, involves studying grps divided by age. Like long., involves examining Ss for a period of time, though shorter time than cross-seq. Cross-seq, control for "cohort" effects, less cost in time and money, less subject drop out.

Front 18

defining feature of true experimental designs?
a. random selection of Ss from the pop.
b. randome assignment of Ss into experimental grps.
c. use of manipulated var.
d. use of non-manipulated var.

Back 18

b. in a true experiment, var. are manipulated and Ss are randomly assigned to tx grps.

Front 19

Galton's concept of regression to the mean best expressed by?
a. individual variation w/in species is unlimited
b. short fathers have tall sons
c. short fathers have taller sons
d. fathers and sons end up having similar heights

Back 19

c. regression to the mean-tendency of extreme oservations to be less extreme upon re-testing/re-observation.

Front 20

which best describes Hawthorne studies at Western Electric plant?
a. changes in temperature and illulmination are likely to result in improved productivity
b. any change in wrkng env. that results in increased attention may result in inc. prod.
c. increases in illum. and decrease in temp. are associated w/better wk perf.
d. decrease in illum. and incr. in temp. are assoc. w/better performance

Back 20

b. Hawthorne effect, tendency of Ss to change their behavior due to attention

Front 21

All are true of multiple baseline designs, except:
a. a treatment is sequentially applied
b. they may serve as a substitute when the ABAB design in unethical
c.may involve studying the same tx for different bx, in diff. stettings, or w/diff Ss
d. they involve admin. and withdrawal of a tx.

Back 21

d. multiple baseline study is a single-subject study involves sequential app of tx across diff. baselines (bx, settings, ind. Ss). Unlike reversal design, does not entail withdrawal of the tx

Front 22

major advantage of case studies:
a. can be used to id var for future research
b. involve the study of only one ind.
c. allow one to draw conclusions about causal relationship btwn two or more var
d. permit generalization of results to other cases

Back 22

a. cannot be used to id causal rel, often used as pilot study to id var. and hypotheses for further study

Front 23

major disadvantage of case studies is that they:
a. can never be used to id var for future
b. invovled the study on only one ind.
c. do not permit conclusions about causal rel btwn var
d. frowned upon by journal editors, professors, etc

Back 23

c. do not permit id of causal rel btwn var.

Front 24

Four classifications of data

Back 24

nominal, ordinal, interval, ratio

Front 25

Nominal

Back 25

unordered categories/who (gender, dx, hair color)

Front 26

ordinal

Back 26

ordered categories/where (ranks, points of a scale)

Front 27

interval

Back 27

continuous data with distance btwn data points is equal across the scale but no absolute 0 (mult/div not possible)- how much (IQ, degrees F)

Front 28

ratio

Back 28

continuous data with equal points with absolute 0 so mult/div possible (how much more)

Front 29

Most observations fall in the middle of the distribution?

Back 29

Normal dist.

Front 30

Most scores fall at the high end of the distribution?

Back 30

negatively skewed

Front 31

Most scores fall at the low end of the distribution?

Back 31

Positively skewed

Front 32

Mode higher than median which is higher than the mean

Back 32

negatively skewed

Front 33

mean is higher than the median which is higher than the mode

Back 33

positively skewed

Front 34

mean, median, mode are all equal

Back 34

normal distribution

Front 35

variance is a measure of?

Back 35

the dispersion/spread/variability of a distribution

Front 36

Standard deviation, a measure of variance, is obtained by?

Back 36

taking the square root of the variance

Front 37

z-score =

Back 37

(X-M)/s.d. where X is raw score, M is mean, s.d. is standard deviation

Front 38

percentile rank is a transformed score that reflects the % of scores falling _____ the corresponding raw score

Back 38

below

Front 39

percentile ranks have a _______ distribution?

Back 39

flat/rectangular; # of scores between 10 and 20 equivalent to number of cases between 80 and 90

Front 40

Almost all transformation of raw scores to percentile ranks would be termed ______ since they would involved a change of original distribution's shape

Back 40

non-linear

Front 41

If you convert raw scores to z-scores you conduct a?
a. linear transformation bc shpe of dist changes
b. linear trans. bc shpe does not change
c. nonlinear trans. bc shape chngs
d. nonlinear trans. bc shpe does not change
b. linear trans.

Back 41

b. when transformed scores retain the same shape, linear transformation.

Front 42

median score

Back 42

middle number of group of scores