Disordered Steps?

"I have been analyzing a mathematics examination with the partial credit model. The test consists of 47 items, each of which can have ratings from 0 to 9. On examining the output, I find some step calibrations are out of order. This occurs because a higher proportion of persons get extreme ratings than get intermediate ratings. Thus the rating of a partially correct performance has a more difficult step calibration than the rating of a perfect performance. Of course, as one goes up the scale, the accumulated category count of all those who got each rating or lower monotonically increases. It would appear that such accumulated counts give a clearer picture of the scale. Do you agree with my reasoning?"

Dear Fred:
You observe that "some step calibrations are out of order". The difficulty of reaching a higher step is less than that of a lower step. This shows that not all your categories are modal. Some categories are never most probable at any ability level. David Andrich sees this "disordering" as a symptom of item or rating scale invalidity.

Geoff Masters points out that step difficulty is not the difficulty of performing the category, but only of observing the category relative to its previous category. Imagine a test in which children are asked to count from 1 to 10. The rating scale is:

does not count up to 1 correctly:0
counts up to 1 thru 6 correctly:1
counts up to 7 correctly:2
counts up to 8 thru 10 correctly:3

Each category represents greater success than the previous category. Observing category 2, however, is unlikely when compared to categories 1 or 3. Consequently step difficulties will be disordered. The step from 2 to 3 will be easier than from 1 to 2. This has nothing to do with the difficulty of the tasks. It is determined entirely by our peculiar specification of the rating scale, a specification that David would criticize.

Geoff suggests that instead of thinking of step difficulties, we think of Rasch-Thurstone thresholds (obtained via step difficulties). You have hit on Thurstone's approach with your "accumulated counts." F. Samejima and P. McCullagh also use models like Thurstone's, but it is not a measurement model because it does not allow parameter separability, i.e., it is not "test-free" nor "sample-free" nor has sufficient statistics.

Thurstone's idea, though, is good for interpretation. The Thurstone threshold is that "median" point on the underlying variable at which categories below are as probable as categories above. These thresholds are always in order along the variable. The interval between thresholds can represent that part of the variable corresponding to a category. As Geoff points out, Rasch-Thurstone thresholds can be calculated from step difficulties. These are the thresholds reported by Rasch programs.

Here are the results for your Item 27 (rescaled in WITS). Category 5 is rarely observed, so it is "hard" (58.8) to get into, and "easy" (47.4) to get out of. The category probability curves show only categories 0, 1, 4, 6, 7 as modal. The thresholds are ordered, but crowded together, because each category corresponds to only a narrow interval on the variable.

----------------------------------
RATING     COUNT   STEP  Rasch-Thurstone THRESHOLD
----------------------------------
   0          15    -        -
   1          11   52.1     49.3
   2           7   53.1     51.3
   3           5   53.0*    52.0
   4           8   49.8*    52.6
   5           2   58.8     53.6
   6           7   47.4*    53.9
   7           5   55.2*    56.8
----------------------------------
* = disordered step calibrations

            CATEGORY PROBABILITIES: MODES
P      ---------------------------------------
R  1.0 +                                     +
O      +                                     +
B      +00                                777+
A      +  00                            77   +
B   .8 +    00                        77     +
I      +      0                     77       +
L      +       00                  7         +
I      +         0               77          +
T   .6 +          0             7            +
Y      +          0            7             +
    .5 +           0          7              +
O      +            0         7              +
F   .4 +            0        7               +
       +             0      *6666            +
R      +              0    *     66          +
E      +        111111*   67       66        +
S   .2 +      11       *4**4         66      +
P      +   111      22***7  4          666   +
O      +111      222 *3***3  44           666+
N      +     2222 ***66*55****5***           +
S   .0 +***************    0*****************+
E      ---------------------------------------
       39    43    47    51    55    59    63

         RASCH-THURSTONE THRESHOLDS: MEDIANS
 p= .5 +0----------1--23456---7--------------+
       ---------------------------------------
       39    43    47    51    55    59    63
                       MEASURE


Disordered Steps?, F Shaw, B Wright, J Linacre … Rasch Measurement Transactions, 1992, 6:2 p. 225


Please help with Standard Dataset 4: Andrich Rating Scale Model



Rasch Publications
Rasch Measurement Transactions (free, online) Rasch Measurement research papers (free, online) Probabilistic Models for Some Intelligence and Attainment Tests, Georg Rasch Applying the Rasch Model 3rd. Ed., Bond & Fox Best Test Design, Wright & Stone
Rating Scale Analysis, Wright & Masters Introduction to Rasch Measurement, E. Smith & R. Smith Introduction to Many-Facet Rasch Measurement, Thomas Eckes Invariant Measurement: Using Rasch Models in the Social, Behavioral, and Health Sciences, George Engelhard, Jr. Statistical Analyses for Language Testers, Rita Green
Rasch Models: Foundations, Recent Developments, and Applications, Fischer & Molenaar Journal of Applied Measurement Rasch models for measurement, David Andrich Constructing Measures, Mark Wilson Rasch Analysis in the Human Sciences, Boone, Stave, Yale
in Spanish: Análisis de Rasch para todos, Agustín Tristán Mediciones, Posicionamientos y Diagnósticos Competitivos, Juan Ramón Oreja Rodríguez

To be emailed about new material on www.rasch.org
please enter your email address here:

I want to Subscribe: & click below
I want to Unsubscribe: & click below

Please set your SPAM filter to accept emails from Rasch.org

www.rasch.org welcomes your comments:

Your email address (if you want us to reply):

 

ForumRasch Measurement Forum to discuss any Rasch-related topic

Go to Top of Page
Go to index of all Rasch Measurement Transactions
AERA members: Join the Rasch Measurement SIG and receive the printed version of RMT
Some back issues of RMT are available as bound volumes
Subscribe to Journal of Applied Measurement

Go to Institute for Objective Measurement Home Page. The Rasch Measurement SIG (AERA) thanks the Institute for Objective Measurement for inviting the publication of Rasch Measurement Transactions on the Institute's website, www.rasch.org.

Coming Rasch-related Events
June 30 - July 29, 2017, Fri.-Fri. On-line workshop: Practical Rasch Measurement - Further Topics (E. Smith, Winsteps), www.statistics.com
July 31 - Aug. 3, 2017, Mon.-Thurs. Joint IMEKO TC1-TC7-TC13 Symposium 2017: Measurement Science challenges in Natural and Social Sciences, Rio de Janeiro, Brazil, imeko-tc7-rio.org.br
Aug. 7-9, 2017, Mon-Wed. In-person workshop and research coloquium: Effect size of family and school indexes in writing competence using TERCE data (C. Pardo, A. Atorressi, Winsteps), Bariloche Argentina. Carlos Pardo, Universidad Catòlica de Colombia
Aug. 7-9, 2017, Mon-Wed. PROMS 2017: Pacific Rim Objective Measurement Symposium, Sabah, Borneo, Malaysia, proms.promsociety.org/2017/
Aug. 10, 2017, Thurs. In-person Winsteps Training Workshop (M. Linacre, Winsteps), Sydney, Australia. www.winsteps.com/sydneyws.htm
Aug. 11 - Sept. 8, 2017, Fri.-Fri. On-line workshop: Many-Facet Rasch Measurement (E. Smith, Facets), www.statistics.com
Aug. 18-21, 2017, Fri.-Mon. IACAT 2017: International Association for Computerized Adaptive Testing, Niigata, Japan, iacat.org
Sept. 15-16, 2017, Fri.-Sat. IOMC 2017: International Outcome Measurement Conference, Chicago, jampress.org/iomc2017.htm
Oct. 13 - Nov. 10, 2017, Fri.-Fri. On-line workshop: Practical Rasch Measurement - Core Topics (E. Smith, Winsteps), www.statistics.com
Jan. 5 - Feb. 2, 2018, Fri.-Fri. On-line workshop: Practical Rasch Measurement - Core Topics (E. Smith, Winsteps), www.statistics.com
Jan. 10-16, 2018, Wed.-Tues. In-person workshop: Advanced Course in Rasch Measurement Theory and the application of RUMM2030, Perth, Australia (D. Andrich), Announcement
Jan. 17-19, 2018, Wed.-Fri. Rasch Conference: Seventh International Conference on Probabilistic Models for Measurement, Matilda Bay Club, Perth, Australia, Website
April 13-17, 2018, Fri.-Tues. AERA, New York, NY, www.aera.net
May 25 - June 22, 2018, Fri.-Fri. On-line workshop: Practical Rasch Measurement - Core Topics (E. Smith, Winsteps), www.statistics.com
June 29 - July 27, 2018, Fri.-Fri. On-line workshop: Practical Rasch Measurement - Further Topics (E. Smith, Winsteps), www.statistics.com
Aug. 10 - Sept. 7, 2018, Fri.-Fri. On-line workshop: Many-Facet Rasch Measurement (E. Smith, Facets), www.statistics.com
Oct. 12 - Nov. 9, 2018, Fri.-Fri. On-line workshop: Practical Rasch Measurement - Core Topics (E. Smith, Winsteps), www.statistics.com
The HTML to add "Coming Rasch-related Events" to your webpage is:
<script type="text/javascript" src="http://www.rasch.org/events.txt"></script>

 

The URL of this page is www.rasch.org/rmt/rmt62i.htm

Website: www.rasch.org/rmt/contents.htm