Polytomous Model Basic Concepts

Basic Concepts

The UCON algorithm for building a Rasch model with dichotomous scores can be extended to more than scores of 0 or 1, i.e. to polytomous scores. These types of models are appropriate for Likert scores and for exams with partial credit.

Notation

Suppose that we have test data with values xsi as defined in Basic Concepts of Rasch Analysis, except that now we will allow scores of 0, 1, 2, …, m where m is any positive integer including values larger than 1. For any ability parameter β and difficulty parameter δ, we define

Probability x = h

whereDifficulty thresholds

Here, the Rasch-Andrich threshold parameter τh is defined as the difficulty of obtaining scoring category h relative to category h −1. Note that the same threshold parameters are used for any subject s and item i. For any item i, the difficulty δi parameter can be viewed as the average of the δi + τh. Let

Gamma

Then for m = 2

and so

Gamma when m = 2

For m = 3,  and P(x = 0) and P(x = 1) are as above and

Note too that –τ1τ2 can be replaced by +τ3.

Iteration

As in the dichotomous case, the ability, difficulty, and threshold parameters are estimated using an iterative approach. In particular, the threshold value τk,t at step t is calculated by

The expected scores xsi,t at step t are calculated as follows:

Expected score at t

The variances at step t > 0 are calculated, as usual, as the sum of squares minus the expected value squared.

Score variance

Ability parameter estimation

The ability parameter for subject s at step t is calculated by

Ability at t

where residuals and variance for subject s at step t > 0 are

Residuals and subject variances

The standard errors of the ability parameters for t > 0 are

Ability standard error

Difficulty parameter estimation

Similarly, the raw difficulty parameter for item i at step t is calculated by

Raw diffculty parameter

where residuals and variance for item i at step t > 0 are

Item residuals and variance

The mean of the raw difficulty parameters is then subtracted from each raw difficulty parameter to obtain the difficulty parameters:

Difficulty parameter

The standard errors of the difficulty parameters for t > 0 are

Difficulty standard error

References

Wright, B.D. & Masters, G.N. (1982) Rating scale analysis. Chicago: MESA Press
https://research.acer.edu.au/measurement/2/

Ataei, S.and Mahmud, Z. (2015) Rasch-Andrich thresholds in engineering students’ attitudes towards learning mathematics
https://www.semanticscholar.org/paper/Rasch-Andrich-Thresholds-in-Engineering-Students%E2%80%99-Ataei-Mahmud/1b60dc7d5a3be98db220d29b957de55e4550d2e2

1 thought on “Polytomous Model Basic Concepts”

Leave a Comment