Stats 1

Front 1

Population

Back 1

Every individual in research scope

Front 2

Sample

Back 2

Selection of population tested

Should reflect population

Front 3

Sample error

Back 3

Difference between sample statistic & population parameters

Front 4

Variable

Back 4

Condition that can change

Independent- changed directly

Dependent- measured change variable

Front 5

Data

Back 5

Multiple observations recorded

Datum= singular

Front 6

Scores

Back 6

Observation through 1 person

Front 7

Data set

Back 7

Collection of data

Front 8

Random sample

Back 8

Every member of population has equal chance to be sampled

Front 9

Representative sample

Back 9

Sample intended to refine and select or adjust for accurate reflection of population

Front 10

Experimental method

Back 10

Only cause and effect method

effect method

*other than very particular case studies

Front 11

Experimental conditions

Back 11

Items that are manipulated

Front 12

Control conditions

Back 12

Remains unchanged

unchanged independent variable

Front 13

Confounding variables

Back 13

Unkown/unseen factors that influence results

Front 14

Extraneous Variables

Back 14

Factors found but not intentionally studied in your experiment or test

Front 15

Third variable effect

Back 15

A type of confounding in which a third variable leads to a mistaken causal relationship between two others

Front 16

Construct

Back 16

Concept that is being observed

Ex: personality

Front 17

Operational definition

Back 17

Builds a method of measure for qualitative items

Front 18

Discrete Variable

Back 18

Hard boundaries between groups

Front 19

Continuous variable

Back 19

Range/spectrum defined as a group without fixed limits

Front 20

4 scales

Back 20

Nominal

Ordinal

Interval

Ratio

Front 21

Nominal

Back 21

Discrete, named categories

Ex: child names

Front 22

Ordinal

Back 22

Descrete, ranked hierarchy

Ex: school seniority classification

Front 23

Interval

Back 23

Continuous, each point reflects space from another point

Category + Hierarchy + Distance

Ex: Credit score

Front 24

Ratio

Back 24

Can't be negative

Interval with fixed origin

Ex: °K

Front 25

Statistics does this with data (3)...

Back 25

Organization- charts/graphs & tables

Summary- descriptive statistics

Interprets- inferential statistics draws conclusions, hypothesis testing

Front 26

Descriptive Statistics measures (4)...

Back 26

Central tendency (average)

Variability (change of score)

Relationships (shared change)

Distortions

Front 27

Mean Median Mode

Back 27

All Averages

Median- middle number of ranked list

Mean- summation of numbers ÷ # of #'s

Mode- most frequent # (can be multiple, or null)

Front 28

Outlier

Back 28

Rare or extreme value

Interaction with mean =heavily

Interaction with median = lightly

Interaction with mode = none

Front 29

N-1 represents

Back 29

Degrees of freedom

Front 30

Covariance

Back 30

Measured common/avg variance

Deviations of score from the mean

E[(x-xbar)(y-ybar)]÷(n-1)

Front 31

Correlation

Back 31

Shared variance ÷ total variance; numerical indicator of magnitude and direction

Correlation from 1 to -1

-1= 1 increases, 1 decreases

0= nothing shared

1= everything shared

Front 32

Coefficients of determination

Back 32

% based common variance

Ratio, negatives don't exist

0 to 1 or 0% to 100%

(n€xy-€x-€y)÷

SQRT[(n€x^2-(€x)^2)][(n€y^2-(€y)^2)]

Front 33

Correlation Types (4)

Back 33

Person Product Moment- most common, interval ratio

Point biserial- 1 variable on interval scale, 1 variable on nominal scale

Spearman Rankorder- variables on ordinal scale

Phi coefficient- 2 variables on dichotomous scale

Front 34

Spurious correlation

Back 34

Illegitimate or false relation between variables due to 3rd Variable Problem or math artefact

Front 35

Chart/graph types

Back 35

Scatterplot- dots

Histogram- connected columns of freq

Freq Polygon- connected line of freq

Pie Chart- % based portions

Ogive- S-curve accumulation of freq

Front 36

Normal distributions... but can have 2 distortions....

Back 36

Are bell shaped

Kurtosis- vertical distortion, height of tails; due to extreme values but doesn't break symmetry

Skew- horizontal distortion

+Skew = above avg outliers Sk>0

-Skew = below avg outliers Sk<0

Front 37

3 types of Kurtosis

Back 37

Mesokurtic- normal K=0

Leptokurtic- tall k>0

Platykurtic- flat k<0

Front 38

Frequency distribution charts are...

Back 38

The most common and basic organization of data

Front 39

2 types of relative distribution

Back 39

Relative- ratio, score freq÷total data

Cumulative- running tally of freq

Front 40

Positive outliers of skew are...

Negative outliers of skew are...

Back 40

Above average and shift the mean more positive along the x axis

Below average and shift mean more negative on the x axis

Front 41

Classic true score model

Back 41

[ X]Observed score = [T]rue score + [E]rror

Front 42

True score and error

Back 42

T: average score one gets on an infinite number of test in parallel format

E: difference between actual test score and True score

Front 43

Random error

Back 43

Affects score through pure chance and is inconsistent

Ex: guessing, distractions

Front 44

Systematic error

Back 44

Effects score because one characteristic has nothing to do with the measured constructs or has serious flaw

Content samples- flaw of test containing wrong subject material

Front 45

Administration error & Score error

Back 45

Can be both Random & Systemic

Admin: usually procedural

Score: difference of raters

Front 46

Test retest reliability

Back 46

Determine tests consistency over time. Scores of the same test compared

Front 47

Parallel reliability

Back 47

Test 1 vs Test 2 with scores correlated (shared variance)

Test 1 --> time--> Test2

Front 48

Inter rater reliability

Back 48

Consistency of individual scores and ratio of agreement

Front 49

Internal consistency reliability

Back 49

Determines consistency of individual items within a test compared to each other

Likert Scale: 1=Hate -> 10=Love

Cronbach's Alpha

Front 50

Validity & 3 types

Back 50

Content. Criterion. Construct.

determines if test measures what you intend to measure

Front 51

Internal & External Validity

Back 51

IV: how much the independent variable changes the dependent variable

EV: application of data/study to the population

Front 52

Content validity

Back 52

Degree of items comprising tests are representative of entire theoretical constructs

Needs a clearly defined content domain and reviewed by subject matter experts

Front 53

Criterion validity

Back 53

What purpose and when is test valid

Test1 --> time --> Test1 --> compare

Post-detective- historical score

Predictive- old score v. Future action

Concurrent- 2 "present" measures compared

Front 54

Construct validity

Back 54

Does it measure an underlying idea

Ex: gravity, personality

Quantifying the qualitative

Front 55

Poponderance

Back 55

51% in favor

Front 56

What is alpha?

Back 56

the likelihood that the true population parameter lies outside the confidence interval (in the extreme tails .025).

Alpha is usually expressed as a proportion.

A=.05