We will employ modern test theory (i.e., Rasch model) to refine the Stroke Rehabilitation Assessment of Movement instrument (STREAM), which will be called the refined version of the STREAM, R-STREAM).

We will employ modern test theory (i.e., Rasch model) to refine the Stroke Rehabilitation Assessment of Movement instrument (STREAM), which will be called the refined version of the STREAM, R-STREAM).

We will devote three years to complete this project. In the first year, the item pool (about 45 items) based on the STREAM will be developed and 12-15 items will be selected from the pool to construct the R-STREAM. A therapist will administer all the items on 300 stroke patients. The other therapist will administer these items on 60 patients randomly selected from the 300 patients to examine interrater reliability of each item. The R-SATREAM (4-5 items for each of the three subscales) will be constructed based on the inter-rater reliability and Rasch analysis of these items, and the opinions of experts.

In the second and third years of this project, the psychometric properties of the R-STREAM will be comprehensively established. We will also compare the psychometric properties of the R-STREAM and the STREAM in stroke patients to determine the psychometric properties and efficiency of the R-STREAM. The R-STREAM and STREAM will be used to longitudinally assess 120 patients (at three time points: at the conception of occupational therapy, one month after therapy, and 6 months after therapy) to compare the concurrent validity, predictive validity and responsiveness of the R-STREAM and STREAM. The test-retest reliability of both scales will be established on the other 60 chronic patients.

It is anticipated that the R-STREAM will have sound psychometric properties, be simple and easy to use, and be able to be transformed to the interval level of measurement. The R-STREAM will best fit the need for simple measures in routine daily clinics. The R-STREAM will fit demands for both in-depth and efficient assessment. Clinicians and researchers both will benefit from these new, scientific measurement tools for planning treatment, giving a prognosis, and assessing outcomes in stroke patients.

Inclusion Criteria:

• patients with stroke ability to follow instructions

Exclusion Criteria:

• patients with other major diseases (e.g., cancer)