ch.6

homogeneous test

A test that measures only one trait or characteristic.

two formulas for estimating internal reliability

KR - 20 and coefficient Alpha

the Spearman Brown formula

Rxx= estimated reliability coefficient of the longer or shorter version of the test n= number of questions and the revised version divided by the number of questions and the original version of the test r= calculated correlation coefficient between the two short forms of the test rxx=nr÷ 1 +(n-1)(r)

test retest method

a test developer gives the same test to the same group of test-takers on two different occasions

test length

adding more questions that measure the same Trader attribute can increase a test reliability test developers can also use the Spearman Brown formula to estimate the number of questions to add to a test so as to increase its reliability to the desired level

Spearman Brown formula

also helpful to test developers who wish to estimate how the reliability Precision of a test would change if the test were made either longer or shorter

coefficient Alpha

and expansion of the KR - 20 formula is used when test questions have arranged a possible answer such as a rating scale may also be used for scales made up of questions with only one right answer

standard error of measurement

as an index of the amount of inconsistency of are expected individuals observed test score the sem is an estimate of how much the individuals observe test score X might differ from the individuals true test score t

intrarater agreement

calculate the consistency of scores for a single scorer A single scorer rates or scores the same thing on more than one occasion

are in the test scoring

can generate air if it is not conducted accurately and according to the direction in the test manual. Although computer scoring is likely to do trees scoring yours it is important to enter the correct scoring scheme into the computer software

errors in the test itself

can generate error by being poorly designed containing trick questions ambiguous questions or poorly written questions or by requiring a reading level higher than the reading level of the test takers

error in the test Administration

can generate your one administrator do not follow instructions for a ministration in the test manual or allow disturbances to occur during the test.

error and test- takers

can you contribute to test error fatigue illness or exposure to test questions before the test can change test questions. in addition test takers you do not provide truthful and honest answers introduce error into their test scores

random error

causes a person's test score to change from one administration of a test the next period defined as a difference between a person's actual score on a test and that person true score

order effects

changes in test scores resulting from the order in which the test were taken

inter-rater agreement

create a rating instrument and have it completed by two judges nominal or ordinal level data. an index of how consistently is scores rate or make decisions

reliability

defined as true score variance divided by the total observe score variance reflects the proportion of the total observed variance in the test scores that is attributable to the variance of true scores

scoring

frequent checks of computation in chess scoring include those made by computers also decrease the chance of scoring errors. Scores he will make qualitative judgments when scoring has such as using a rating scale must receive training together to collaborate their judgments and responses

reliable test

is one we can trust a measure each person in approximately the same way every time it is used

split half method

is to divide the test into halves and then compare the set of individual test scores on the first half for the second of individual test scores on the second half.

heterogeneous test

measuring more than one trait or characteristic estimates of internal consistency are likely to be lower

systematic error

on a single source of error always increases or decreases the true score by the same amount

test Administration

proper test Administration affects the reliability estimate in three ways The test takers here the same instructions and take the test under the same physical conditions each time. Treating all test-takers in in the same way decreases are that arises from creating differences in the way individuals respond. Consistency between two administration's decrease error that arises when testing conditions differ. effective testing practices decrease the chance the test-taker score will be contaminated what are due to for testing conditions or poor instructions

formula for random error

test score equals true score plus random error or x = t + e

scorer reliability or interscorer agreement

the amount of consistency among scores judgments becomes an important consideration for Tessa require Decisions by the administrator or score

true score

the first part is a measure of the amount of the attribute that test measures. represents the score that would be obtained at the end of vigil took a test in infinite number of times and then average score across all the tests were computed

test retest interval

the longer the interval between administrations of test the lower the reliability coefficient is likely to be. A long and a roll between test administrators provide more opportunity for testicular to change in terms of the factor being measured

random error E

the second part of and obtained test score consists of random errors that occur at any time a person takes test.

generalizability theory

this Theory analyzes the many causes of inconsistency are in test scores and seeks to separate sources of systematic error from random error

spearmint and brown formula

this formula it is designed to estimate what the reliability coefficient would be if the test had not been cut in half but instead where there original length we typically use this formula when adjusting reliability coefficient arrive by correlating two halves of one test

test retest method

this method is appropriate only when test takers are not likely to learn something for the first time they take the test that can affect their scores on the second Administration or when the interval between the two administrations is long enough to prevent practice effects

alternate or parallel forms

to overcome problems such as practice effect psychologists often give two forms of the same test designed to be as much alike as possible to the same people. This strategy requires the test developer to create two forms of the test I referred to as this

KR 20

use for the skills that have questions that are scored either right or wrong such as true or false and multiple choice questions

practice effects

when testicles benefit from taking the test the first time which enables them to solve problems more quickly in the correctly the second time

intrascorer reliability

whether each clinician was consistent in the way he or she assigns scores from test to test