Evaluation of Formal Assessment Instrument

Kathman 1

Justin Kathman

Ed 314

Professor Deterding

10/2/14

Evaluation of Formal Assessment Instrument

Title: Test of Early Reading Ability (TERA 3)

Author: D. Kim Reid, Wayne P. Hresko and Donald D. Hammill

Publisher: Pro-Ed (Austin, Texas)

Description:

General Purpose: The TERA3 is a norm referenced test intended for children between the ages of 3 years and 6 months, (3-6) and 8 years 6 months, (8-6). The TERA3 is made up of three subtests including: Alphabet, Conventions, and Reading. Subtest 1 contains 29 items, subtest 2 contains 21 items, and subtest 3 contains 30 items. The student earns 1 point for each item passed, and incorrect items are marked with a 0. This test was created for the following purposes: to identify children who are significantly below their peers in reading development, to identify strengths and weaknesses of individual children, to document a child’s progress as a consequence of early reading intervention programs, to serve as a measure in research, and to accompany other assessment techniques.

Materials:All materials are provided-examiner’s manual, picture book forms A and B, and examiner’s record booklet forms A and B.

Alternate forms: There are two forms to this test, forms A and B. Both forms contain the same subtests and all three of the subtests contain the same number numbered items in each.

Administration:

Age Ranges: 3- 8 or grades preschool to second grade.

Administration Time and Scoring Time: Approximately 15 to 45 minutes. This test is not a timed test so there are no imposed time limits on the children being tested. Things to consider though are children below the age of 5 may need a break every 10 minutes. Also, the examiner should encourage the student to progress rapidly to avoid procrastination. Scoring is not lengthy: the raw score is obtained by counting up the number of correct items on a subtest and then charts are used to convert raw scores into standard scores.

Types of Scores Reported: The TERA reports 6 types of scores: raw score, standard score, reading quotient for the composite, percentile, age equivalence, and grade equivalence. The raw score is calculated for each subtest by adding up the correct number of items and then converted into the stanines and percentiles using the normative table. Stanines are based on a distribution with a mean of 10 and a standard deviation of 3. The reading quotient is derived by adding up the standard subtest scores and then converting the sum to a composite score in a chart. Since the quotient is a standard score, it has a mean of 100 and a standard deviation of 15. Age and grade equivalents are derived by calculating the average normative group score at each 1-month interval from age 3 years 6 months through age 8 years 6 months.

Standard Error of Measurement and Confidence Intervals: The standard error of measurement is the same for both forms. The standard error of measurement for the subtests is 1 while the standard error of measurement for the reading quotient is 3. The SEM can be reported with a 68% confidence band to obtain the true score. The test also reports that the SEM could be calculated with a 95% confidence band= (1.96x3), or with a 99% confidence band= (2.58x3).

Norming Procedures:

Sample Size: 875 children residing in 22 states were tested.

Sampling Procedure: Over 875 students from 22 different states residing in four different geographic regions, northeast, south, mid-west, and west, where tested. At each geographic region, usually a major city within the region, teachers were asked to select students no matter their ability level. Then participants where tested primarily at day care centers, preschool facilities, public schools, and private schools.

Characteristics of the Sample: The test sample included an even number of males to females. Characteristics also reported race, residence, ethnicity, geographical regions, family income, educational attainment of parents, and disability status. The normative group was broken into age groups, showing the number of participants for each age. Number of participants ranged from 89 to 231 individuals for each age 3-8 years.

Date of Norms: Between February of 1999 and April 2000.

Reliability: There is an alternate form of this test. Both tests are parallel and the means and standard deviations for forms A and B are similar at every age interval. All the correlation coefficients depicting the relationship between the forms exceed .80. This concludes that the TERA3 evidences high reliability between the two forms. Reliability was reported in regards of three sources of error variance: content sampling (internal consistency), time sampling, (test retest), and scorer differences.

Internal consistency: This was reported for each age interval used in the TERA 3 and a reliability coefficient was reported for each age (3-8). The range of coefficients for each scale, with the decimal omitted.

Alphabet A84-94MeaningA80-94

B84-95B84-95

Conventions A79-88

B75-89

Test Retest Reliability: Professionals gave the test to two different groups of normally achieving students. One group was made up of 30 students ages 4-6 attending public school in OkimosMichigan. The other group was made up of 34 normally achieving students ages 7-9 from Austin Texas. Both forms of the tests were administered twice to the sample with an intervening time of 2 weeks. The following reliability coefficients were reported for each scale. Decimals omitted.

Total Sample

AlphabetA96ConventionsA92MeaningsA92

B96B93B94

Validity

Content Validity: Content validity is reported in the following ways. 1) By reviews of existing research, commercial and non-commercial curriculum materials, and popular tests show that the TERA3 items reflect the current state of knowledge. 2) The professionals compared there work to the work of others to show that the TERA compares favorably to that of other reading experts. 3) By having experts examine the items within the TERA, the experts indicated if they liked the placement of each item in the three subtests. 4) How conventional item analysis procedures validated each item during the developmental stages of test construction. 5) The results from a differential item functioning analysis show a lack of bias in the test items.

Criterion Prediction Validity:Criterion prediction validity evidence for this test was generated by correlating scores obtained on a new test with scores on other tests of the same or similar abilities. To test for criterion prediction validity, they tested the correlation between the TERA3 and the TERA2. The test tests where compared by correlating Forms A and B (TERA3) subtest and composite raw scores with the composite raw scores of the previous edition.

Construct Identification Validity: The construct identification validity of a test is the extent to which the test may be said to measure a theoretical construct or traits. There are seven basic constructs that underline the TERA3 and other testable premises. 1) Performance on the TERA3 should be strongly correlated to chronological age. 2) Results should differentiate between groups of people known to be average and those to be known as below average in reading ability. 3) Since all the tests measure reading, they should correlate significantly with each other. 4) TERA3 should correlate with measures of school achievement. 5) Scores should relate to tests of intelligence, especially those that measure verbal abilities. 6) A confirmatory factor analysis should support the relationship of the subtests to constructs inherent in the test model. 7)Items of each subtest should be highly correlated with the total score of that subtest.

Classroom Uses

Suggested by Authors:The purpose of the TERA3 is to identify children who are struggling in reading development, identify strengths and weaknesses of an individual student, helps in documentation, and it accompanies other assessment techniques. General educators can use this information to help guide early intervention programs and the TERA3 can help find strengths and weaknesses in your students reading ability and can help the early intervention process. The TERA3 can also be used alongside a variety of other comprehensive assessments, like diagnostic interventions, and portfolio collections of the child’s work. One thing to note is that the TERA3 should not be used to guide your lesson plans. The TERA3 should be used as a guide to areas that need to be explored further.

My Opinion: The TERA3 would be extremely useful in finding strengths and weaknesses in students. Also I like the fact that the TERA3 could be used with many assessments to help areas of need in reading instruction.

Desirable/Undesirable features

Desirable Features / Undesirable Features

All materials provided
Useful in pre-referral.
Useful in contribution to MDT process.
Reliability was generally solid.
Validity was also solid.
Quick assessment. Approximately 15 to 45 minutes to administer.
Easy scoring.
Information and raw scores can help early intervention process.
3 subtests.
Easy to Administer.

Test-Retest reliability was done over a two week span of time. Needs more of a span.