Test Review: Transition Planning Inventory−Second Edition By J. R.,Patton & G. M. Clark

Materials

The TPI-2 kit includes the Administration and Resource Guide booklet, Informal Assessments for Transition Planning–Second Edition booklet, Transition Instruction Guide: Standards-Based Activities booklet, 25 Profile and Further Assessment Recommendations Forms, 25 School Rating Forms, 25 Home Rating Forms, 25 Student Rating Forms, 25 Preferences and Interests—Basic Forms, 25 Preferences and Interests—Advanced Forms, and the two Resource CDs (i.e., Informal Assessment for Transition-Second Edition Reproducible Forms CD and TPI-2 Administration and Resources Guide: CD Resources). The forms are user friendly. The color-coded forms allow for easy access when compiling them.

Administration

TPI-2 includes a set of questionnaires that can be completed with the support of a school-staff using individual or group format. The staff is advised to follow a script to administer these forms, but the script can be modified as needed. It also provides for alternative administrations (self-administration, oral, or guided). The staff administering the test can be a teacher or an individual with a background or proper coursework in educational measurement. The school-staff will be responsible for handing out the rating scales and compiling the responses into the Profile and Further Assessment Recommendation Forms. The individuals responsible for filling out the questionnaires typically include the student (Student Rating Form), the parent or guardian (Home Rating Form), and the IEP teacher(s) or case manager (School Rating Form). The total administration time is 15 to 30 min, and the order for administration is as listed below:

The targeted population is students (ages = 4-22) who are receiving SPED services and have reached the age that mandates transition planning, which varies by state.

Scoring System

The qualitative data can be processed by hand or computer. There is no traditional scoring, rather the answers are transferred and visually displayed on a single form (Profile and Further Assessment Recommendations Form). There are no standard scores, age/grade equivalents, percentiles, or any normative data. The hope is to inform decisions around transition using a qualitative approach. Specifically, responses are captured using a 57-item Likert-type scale that ranges from zero (strongly disagree) to five (strongly agree) with not appropriate and do not know as alternative responses. The statements that are rated a three or above are regarded as (the student) having an acceptable level of competency, and statements rated 2 or below show a “lack of competency” and an area that needs further planning. Discrepancies across responders are determined when a rater indicates that the student has an acceptable skill (i.e., score ≥ 3), and another rater answering the same statement provides a score of 2 or less. The goal is to bring stakeholders together and discuss the difference of opinion in rating the student’s abilities and ideally come to an agreement, which consists of stakeholders assigning similar scores to a given statement.

Test Construction

As part of the construction, a survey was sent asking about usefulness and clarity of the TPI among the initial individuals who purchased the TPI (i.e., every 30th purchaser of the original version). The feedback was used to select a preliminary pool of items from the original version, add new items, and/or revise items. As a next step, the preliminary version was sent to a group of 64 professionals (representing 19 states) with background, training, and experience in transition services. The general agreement was that the instrument was comprehensive and targeted knowledge and skills relevant to transitioning. Their feedback also included suggestions for new items or clarifying/deleting items, resulting in a 20% increase of items in the TPI-2.

There is no evidence confirming or disconfirming gender or racial bias; however, from the sample, there appears to be an underrepresentation of Hispanic (n = 62) and Asian (n = 8) students with disabilities, and there was no mention of whether Latino students were included in their sample.

Standardization Sample

The TPI-2 was standardized in 2010 with a sample of 535 students receiving SPED services. Although the TPI-2 is marketed for individuals aged 4 to 22, the ages of the standardization were listed as 9 to 25 and with a mostly male sample (64%). Race was distributed as follows: 79% White, 17% Black/African American, 1% Asian American, 2% two or more races, and 1% Other; 12% of the sample identified as having a Hispanic ethnicity. The sample was collected from 14 states. The number of disabilities reported among the sample was mostly one disability (i.e., 83%), followed by two disabilities (12%), three disabilities (3%), and 2% reported more than three disabilities. The sample did not include individuals with language impairments. The disabilities most represented are listed in order as follows: learning disabilities (n = 267), intellectual disability (n = 111), otherwise disabled (n = 81), emotional disturbance or behavioral disorders (n = 41), and autistic (n = 35). This sample was used for the sampling reliability analyses; however, smaller subsamples were used for test–retest and criterion validity analyses.

Reliability

The reliability of the TPI-2 was looked at in terms of content sampling, time sampling, and interrater reliability. Content sampling was used to measure internal consistency or the degree to which the statements in each broad category correlated with one another, with alpha coefficients of .80 or greater showing adequate reliability. Internal consistency reliability was calculated for all raters and the four largest disability classification groups. Content sampling reliability for the broad categories showed coefficient alphas of greater than .80, meaning that the statements contained in the broad categories were highly related to each other. Content sampling was also assessed among the statements on each of the 11 narrow planning areas. When looking at the reliability coefficient for statements on these smaller categories, 50% of the statements were reliable (i.e., ≥.80). TPI-2 explained this finding by arguing that the number of statements on each of the 11 planning areas only had four to eight statements on each category capturing a very limited amount of variance.

Time sampling consisted of a 7-to-10-day interval with a subsample of 55 cases (i.e., teachers who volunteered). Results showed that six (or 67%) out of the nine broad areas coefficients were reliable (i.e., ≥.80), and 52% of the 33 coefficients among the planning subcategories were reliable. Teachers’ responses were the most reliable (100%). Parents’ responses were reliable in four out of the 11 subcategories, and student’s responses were reliable on two of the 11 subcategories. Of note, the lowest test–retest reliability coefficient was found among students’ rating of the working broad-area and specifically the “Career Choice and Planning” subcategory. It is fair to mention that it is not uncommon for youngsters to have fluctuating ideas of their future career choices, and this lack of reliability could be capturing a normal developmental process of decision making at that age.

Finally, interrater reliability was used to measure the degree of agreement between raters with regard to the student’s strengths and/or weaknesses in areas that involve postsecondary (working, learning, and living) transitions. Two individuals, independent to the test development, randomly reviewed 10 cases. For each case, there were three different raters: the student, the caregiver, and a teacher. The overall percentage of interrater reliability was 99%. Most importantly, findings also documented the potential perils of only having one rater (i.e., the student) when planning for after school transition. To have reliable results, this instrument should be used with at least two raters and/or ideally all three raters’ feedback.

Validity

Validity of the TPI-2 was captured using a validity model proposed by Linn and Gronlund (2000). This model proposed that validity should be looked at in terms of the interpretation of the results rather than the instrument itself. The TPI-2 used this model of validity to justify not assessing for predictive validity in their standardization design, citing that this instrument is a domain-referenced rating scale and therefore not intended for predicting future performance. The research design concentrated on demonstrating the validity of the TPI-2 by looking at content and criterion validity only.

The content validity assessment was done during field testing by looking at the degree to which the content of the TPI-2 was age-appropriate and relevant (items are consistent with current knowledge in the area of transition planning) and by assessing criterion validity. The relevance of the TPI-2 items was assessed via survey and feedback from field-test participants and individuals deemed experts. The feedback from these two groups documented that a majority (87%-94%) of responders agreed that the instrument was responsive to the IEP planning process under IDEA and that it covered the basic considerations for transition planning.

Criterion Validity was measured using the concurrent validity method, which compares the TPI-2 with another instrument (i.e., Enderle-Severson Transition Rating Scale, Form J-Revised) measuring a similar construct. The results documented that the broad areas of working, learning, and living showed moderate to very large inter-correlation among these two instruments.