Mathews Eval ch 2 test 1

Standardization

the development of invariable procedures of measurement and interpretation of measured values

1. careful selection of test item

2. administration of the test items to representative sample drawn from a defined population

3. statistical analysis of results

4. establishment of age-based norms

5. development of instructions and response scoring procedures

The standardization process includes

1. self-contained

2. stimulus materials

3. record forms

4. examiner’s manual

Standardized tests include

Self-contained

all items needed to effectively administer the assessment should be included.

Stimulus materials

any materials used in the tasks for measurement

Record forms

where you record the responses from the client and score the test

Examiner’s manual

Used by the examiner to obtain scores for the standardized test

Norm

performance measure of a normative group on a tested skill

normative group

a representative sample of individuals of the same age and possibly the same sex as the client

Norm-referenced test

compares a client’s performance to the performance of another group of individuals called the normative group

Size of the sample

Ethnocultural and socioeconomic variables

typical and atypical samples (typical sample will be chosen to differentiate or identify the disorder)

Geographic distribution

Descriptive statistics

Normative Sampling should consider

Bell curve

typical normal distribution of scores results in a bell curve with the ranges of scores plotted on the horizontal access and the number of participants receiving a particular score on the vertical access

Mean

statistical average of individual performance

Standard deviation

individual’s score may not correspond to the mean; it might deviate and the extent to which it does is called the standard deviation

Raw Score

is the initial score given based on the number of correct responses to test items

Scales or derived scores

the scores derived from the raw score that allows different individuals to be compared to determine strengths and limitations

Percentiles or percentile scales

represent the percentage of individuals in the standardization sample scoring at or below a given raw score

z-scores

linearly derived standard scores based on assumption that the mean is zero and the standard deviation is 1

T-score

normalized standard scores

cut to fit a normal probability curve

mean is 50 and SD is 10

Stanine

based on a range of 9 numbers with a percentage of tested individuals connected to each number

Basal scores

basal score is the entry level or initial score

Ceiling score

is the test terminating score after a number of incorrect responses

1. test manual

2. Admin and scoring procedures

3. Stimulus items

4. normative sample

5. statistical analysis

6. validity and reliability

Test components

Reliability

the consistency across repeated measures of the same phenomenon with the same instrument

A ___ test yields the same or similar score when an individual is repeatedly tested by a single examiner or by different examiners

1. Interobserver Reliability

2. Intraobserver/Test-retest Reliability

3. Alternate or Parallel Form Reliability

4. Split-half reliability

Types of reliability

Interobserver reliability

consistency of test scores recorded by two or more examiners administering the same test to the same individuals

Intraobserver

consistency of scores the same individuals obtain when the same examiner readministers the test or repeats naturalistic observation

Alternate/parallel form

the consistency of measures when two forms of the same test are administered to the same persons

Split-half

a measure of internal consistency. correlates scores from one half of the exam and compares them to the other. even/odd can be harder to test.

Validity

the extent to which a test measures what it is constructed to measure. it reflects the truthfulness of that measure

Content Validity

Construct Validity

Criterion Validity

Types of Validity

Content Validity

when an expert exams the test items and concludes that the items assess the purported skill

Construct validity

the degree to which a test’s measures are consistent with a known theoretical construct. A construct is a theoretical statement supported by prior empirical evidence

Criterion validity

the assurance that a test measures what it purports to measure because it is correlated with an external variable called a criterion

1. detailed and comprehensive test manual

2. large and diverse normative sample

3. samples skills adequately

4. recently revised and updated normative data

5. strong reliability and validity

6. appropriate stimulus items

7. well-trained to administer

8. yield useful diagnostic information

Criterion for Selection of Standardized test