Midterm Easy Flashcards
Frequency Polygon/histogram
Both ways to plot out test scores and frequency
X axis = scores
Y axis= number of people who get scores
T-Scores
this for some reason exists because people don’t like seeing negative numbers
In this scoring, the mean is always 50, and the SD is always 10
To get this score, multiply z score by ten and add 50
Negative Skew
Scores pile up on the high end
Mean < Median > Mode
(Questions too easy?)
Variability
The extend of individual differences around the central tendency
Range
Interquartile range
Standard Deviation
Likert scaling
Frequently used summative rating scale
Typically five response options on a continuum
1=strongly disagree, all the way to 5 =strongly agree,
Has a neutral
Norms
Understanding a person’s test scores in relation to others that have taken same test
Nominal
Replaces names with numbers
Names belong in one category
1=AA 2=White 3=Latinx etc
Used for purposes of identification
Cronbach’s coefficient alpha
Reliability test for internal consistency
For continuous responses
1=never 5=always
Forms of reliability are also considered internal consistency methods that attempt to estimate the reliability of all the possible split-half
individuals who are producing depressed scores should be responding to test items in a manner that indicates depression, and individuals who are producing scores that indicate they are not depressed should be responding to test items in a manner that do not indicate depression.
Interval
Equal intervals between each number
No absolute zero
Used for identity, rank, equality of units
Fahrenheit/Celsius degrees, same number means different thing
Content validity
Adequacy of representation of content Is everything included? Is anything missing? Under represented Irrelevant
Validity
Agreement btw test score and quality it is believed to measure.
Does the test assess what it is supposed to?
Provide proof for verification of validity
To a degree, not a yes or no
Paired comparison
Pairs of stimuli
Select one based on a rule
Which word appeals more? This or that
Where did assessments originate?
China, to place civil servants
Semantic differential
Anchors either side of continuum
Contrast two extremes
Anxious__1_2-3 calm
woodworth
Created first objective assessment of personalities for troop placement
Types of validity evidence
Content
Criterion
Construct
Ratio
Same as interval scale but there is true zero
Annual income, number of arrests
Errors in Measurement
Test construction
Test administration
Test taker
Examiner
Convergent validity
His correlation btw test and criterion thought to measure same thing
Correlates high with other tests going for the same thing
Has this and discrimination, test has construct validity
True score
Hypothetical score including errors of measurement
Gutmman
Response to categories
The one where the stronger item is selected out of a list of progressively stronger statements from weaker to strong, whichever level of statement you choose, all weaker statements are agreed to as well.
Test retest
Reliability test Test once, then again after time States vs traits States tend to change over time so use shorter interval Traits stay stable so longer interval
Standard error of measurement
Tells us how much a score varies from the true score
Keep in mind the normal distribution with the middle 68% confidence
Predictive
Validity evidence
Scores used forecast for some future behavior
Aptitude tests
Get scores, then look for predictions at later point
Range
Subtract lowest score from highest
Criterion-Refereced
Test interpretation that compares scores with absolute score.
Driving test
Certification test
Mastery test
Construct validity
Theoretical or scientifically informed idea that describes or explains some behavior
Intelligence, self esteem, motivation depression
Binet
Created the Simon-Binet scale that eventually turns to the IQ test that we know
Test originally to assess higher mental processes by evaluating test takers task accomplishment
Coefficient of determination
Shows how much of variation in one measure is accounted for by knowing the value of the other measure
Used for validity
Squared the person r coefficient, gives us the percentage of relation between two assessments,