CLP 4433 TEST NO.1

measurement instrument that attempts to measure personal attributes in understanding/describing human behavior

• 1. sample of behavior
• 2. obtained under standardized conditions
• 3. established rules for scoring

performance, behavioral observations, self-reports

describe subject behavior in response to test in quantitative/numerical terms

2 people with the same set of scoring rules applied will arrive at the same score

relies on examiner judgement

test results influence individual's actions

• benefit - feedback, guides for decision making
• concerns - testing can have undue impact

results used by organizations for placement purposes

• benefits - can be used for employment
• concerns - accuracy and fairness questionable

selection of individuals from a group of test takers

ex. employment test

based on single person on test based on demonstration of mastery

ex. promotion testing

• 1. impact of ability testing
• 2. fair use of tests and outcomes
• 3. invasion of privacy and confidentiality

difference in scores that aren't attributed to the construct

group mean difference after removal of potential bias

reflect REAL differences on the construct

• 1. there are obstacles for some groups (ex. gender, physically/mentally impaired)
• 2. they are the SOLE basis for decisions
• 3. poor performance = harsh consequences

multiple assessment procedures, hurdle system, compensatory system

multi-stage decision model with more intensive screening for some groups

low performance in one area can be made up for by high performance in another

requires "reasonable accommodation" when testing/employing disabled individuals

individuals directly affected by decisions of tests

most common: test-takers, test-developers

fairness from institutional perspective

comparative decisions

success of test and long-term results

personal consequences

absolute decisions

cost-effectiveness and short-term results

Nominal, Ordinal, Interval, Ratio

numbers or labels to identify objects, persons, groups

ex. football player jersey numbers

rank order - order of numbers correspond to order of objects (not mathematically manipulated)

ex. school percentiles

uses equal units for differences in which objects are ordered

ex. temperature

meaningful differences using an absolute zero

can be mathematically manipulated

ex. money

a statistical measure to determine a single score that defines the center of distribution

the most frequently occurring value in a data set (can be more than one)

symmetric - mirror images

skewed - asymmetrical; median used for measure of central tendency

deviation score = mean - each person's score

the extent to which scores in a distribution differ from one another

determined by the amount that values differ from the mean

types of measures: range, variance, standard deviation

indicates how far individual values are from the expected value (mean)

average of the SQUARED deviation scores

Large - individuals differ in score and from each other

Small - individuals are similar

• 1. COMPUTE the MEAN
• 2. SUBTRACT the mean from every observed value
• 3. SQUARE the values and SUM them
• 4. DIVIDE the sum by the NUMBER OF VALUES

square root of the variance

the extent to which two variables are systematically related, quantitative estimates