Stats Exam 1

Descriptive Stats

Describe and Organize characteristics of data

Inferential Stats

Inferences on samples/population

Data set

Collection of data

Sample

A smaller group within a population (n)

Population

The entire pool of data you want to analyze (N)

Mean

1. average

2. organize data, add all #’s, divide by n

Median

1. organize data, middle # is the answer

2. If there’s two numbers in the middle find mean of two #’s

Outliers

Usually far away scores

Weighted Mean

when there is a limit number of values that occurs many times

When to use Mean

no outliers

When to use Median

When there’s outliers

Percentile Points

percentage of scores below a particular score

Levels of Measurement

Nomial

Ordinal

Interval

Ratio

Nominal

Categories

Ordinal

Categories that are rank

Interval

Range of numbers but zero doesn’t mean that there’s nothing

Ratio

Range of numbers with a true zero so zero actually means nothing

Ex,) height & weight

Variable

Thing being observed/measured and has more than one outcome

Value

score, category, number

Independent Variable

what is manipulated/ impacted by another variable

Dependent variable

outcome measurement

Measures of Variability

Exclusive, Inclusive, Standard Deviation

Variability

How scores differ from eachother

Exclusive Range

Highest value - Lowest Value

Standard Deviation

S, uses a square root

Table: X, Mean, X-Mean, (X-Mean)², SQUARE ROOT ∑(x-mean)²/n-1

Variance

S², Doesn’t not use square root

Table: X, Mean, X-Mean, (X-Mean)²,∑(x-mean)²/n-1

Steps for Frequency table

1. Class Interval of 10

2. Find Range ( H-L)

3. Find Width ( Range/Class Interval) ROUND UP

4. multiple of width ≤ lowest value

5. start from

6. Bottom

Table for Frequency Table

Table: Title+ Labels C.I (X-axis), Tally, Frequency(Y-axis)

Discrete Value

Dollars & Midpoint

Continuous Value

Cents, on the vertical line

Cumulative Frequency

Running Total, answer=n=100%

Cumulative Percentage

Running total percentage, answer=n=100%

Rules for Histogram

title, label x&y values, discrete or continuous,

Line Graph

Display trends

Pie Chart

Display proportions

Rules for Pie Chart

1. Title, add sample size

2. Legend/Label

3. start at 12o’clock

4. create clockwise, descending order

5. give % in each slice

Table for Pie Chart

1. category

2. frequency

3. Proportion (p=f/n) don’t round answer

4. Size of Slice ( p x 360 )

5. Percent (p x 100)

Bar graph

Comapre Categories

Frequency Polygon

title, label values, discrete or continuous, connect dots at top of Midpoint, ground them

when does SD = 0

all scores are equal

Ogive

Cumulative frequency

correlation coefficient

relationship between X & Y, highlighting their strength and direction

Pearson Product-Moment Correlation Coefficient

measure strength and direction

Positive/Direct Correlation

+1.00

both variables affect one another

X increases Y increases or X decreases Y decreases

Negative/Indirect Correlation

-1.00

variables change in opposite directions

X increases Y decreases or X decreases Y increases

Zero Correlation

Both variables don’t affect each other

Pearson Product Table

1. Label 5 columns N, X, X², Y², XY

2. list X & Y

3. find ∑X, ∑Y, ∑X², ∑Y² of all columns

4. 
   

Pearson Product Formula

this is r

Coefficient of Determination

r²

explains % of variablity

Coefficient of Nondetermination

k² = 1 - r²

does not explain Variability

Reliability

Consistency

Observed/Total score

Actual Score you get

O = True Score + Errors / True score

true score

just theoretical

Error Score

Silly mistakes or lucky guesses

E = total score - True Score

Test - Retest

1 type of variability

measures consistency overtime / stability

Parallel Forms

2nd type of variability

Measures Equivalence,

ex.) is Test A = Test B?

(a) greater than 0.6

Internal Consistency

3rd Type of Variability

Measures Consistency

How consistently the items within test A & B measure the material learned.

Cronbach’s Alpha

Inter-Rater

Measure of agreement

amount of time people agree on something, doesn’t mean it’s true

Validity

Accuracy

Content Validity

Does it accurately measure what we need to know\

more logical

Criterion Validity

Does the test tell you what the other does

has two parts

Concurrent Validity

“current”

how accurate does the outcome of test correlate to a similar test right now

Predictive Validity

“Predict - future”

how accurate does the outcome of test correlate to a similar test taken in the future

Construct validity

how well a test reflects behavior