Eppp Statistics Flashcards

Front 1

Nominal

Back 1

unordered categories

Front 2

Ordinal

Back 2

ordered amounts, rankings, likert

Front 3

Interval

Back 3

equidistant amts with no absolute zero point; (can add/subtract values but not multiply/divide)

Front 4

Ratio

Back 4

equidistant amts with absolute zero point; (can add/subtract/multiply/divide)

Front 5

Frequency distribution

Back 5

# cases within a given category or score range

Front 6

Variance

Back 6

the standard deviation squared (squared differences of each observation from the mean)

Front 7

Std Deviation

Back 7

square root of the variance; can be used to calculate the % of scores that will fall within a given range or at/below a given cutoff

Front 8

Skewed distribution: Positvely skewed: most scores are in the low range (hard test) Negatively skewed: most scores are in the high range (easy test)

Back 8

Positive: running a mile in 5 minutes Negative: Walking a mile in 1 hour

Front 9

linear transformation

Back 9

converting a raw score to a std score does not change the distribution shape (curve)

Front 10

nonlinear transformation

Back 10

changes the distribution's shape when raw scores are converted

Front 11

Stanine

Back 11

divides the distribution into 9 intervals (mean=5, sd=2)

Front 12

Percentile

Back 12

% total scores falling below the given score

Front 13

Statistic vs. parameter

Back 13

Statistic=sample value Parameter=pop.value

Front 14

Sampling error

Back 14

inaccuracy in estimating the pop mean from a sample mean

Front 15

Std error of the mean

Back 15

the expected deviation bt the sample mean and the true pop. mean.

Front 16

True Experimental

Back 16

Random assignment to groups that receive differential levels of a manipulated variable; looking for causality. Greatest internal validity.

Front 17

Quasi-Experiment

Back 17

Assignment is not random but group variables are still manipulated; looking for causality

Front 18

Correlational

Back 18

Measures the degree of relationship between 2+ variables for the same group. Variables are not manipulated, no causal relationship assumed. Goal is to use 1 variable to predict status on a 2nd variable.

Front 19

Developmental

Back 19

Longitudinal, cross-sectional, cross-sequential (Ss of different age groups are studied over a short period of time); Correlational over time; no causal relationship.

Front 20

Time-Series Interrupted Time Series

Back 20

DV is measures several times at regular intervals, both before and after a tx. Often involves only 1 grp and 1 tx (controlling for threats to internal validity, except history); type of Quasi-Experimental design, uses a time-series analysis

Front 21

Single Subject

Back 21

1 Subject, at least 1 baseline and 1 tx phase. DV is measured several times during both phases: AB, reversal (ABAB) and multiple baselines (ABC)

Front 22

Qualitative (Descriptive)

Back 22

Data collection to arrive at a theory of "how things are."

Front 23

Scientific research

Back 23

Data collection to test an existing theory

Front 24

Factorial experimental design

Back 24

Every level of each IV is compared with every level of all other IVs

Front 25

Latin Square

Back 25

experimental design in which groups receive txs in every different order

Front 26

Solomon Four-Group

Back 26

Experimental design to reduce re-retest problems; four comparison groups are used