Email discussion lists for the UK Education and Research communities

## WEB-ASSISTED-ASSESSMENT@JISCMAIL.AC.UK

#### View:

 Message: [ First | Previous | Next | Last ] By Topic: [ First | Previous | Next | Last ] By Author: [ First | Previous | Next | Last ] Font: Proportional Font

#### Options

Subject:

Re: facility and discrimination

From:

This list, hosted by the CASTLE project, is for those interested in conduct" <[log in to unmask]>

Date:

Fri, 23 Mar 2001 12:46:31 +0000

Content-Type:

TEXT/PLAIN

Parts/Attachments:

 TEXT/PLAIN (49 lines)
 ```At Aberdeen, we had a program written for item analysis back in 1973 by J Houston and J Mitchell, with advice from people then working with Alex Johnstone. The program has been modified for use with omr forms, but we still use it now. This is what the manual says about facility and discrimination: ITEM ANALYSIS After scoring and sorting into merit order the lists are searched and the following quantities are calculated for each question in the exam: 1. The facility value, i.e. the fraction of the class with the correct answer. 2. The discrimination value, computed as the point biserial correlation coefficient between response and total scores, gives a measure of how effective the question is in discriminating between high scorers and low scorers. Note: The point biserial correlation coefficient for a particular question is calculated as: ( Mr - Mt ) * sqrt( Nr / Nw ) / SD where:  Mr = Mean score of candidates giving CORRECT answer.  Mt = Mean score of the whole class.  Nr = Number of candidates giving CORRECT answer.  Nw = Number of candidates giving WRONG answer.  SD = Standard deviation of scores for whole class:          sqrt( N * S1 - S2 ) / N      where: N = Number of candidates in class.          S1 = Sum of squares of scores.          S2 = Sum of scores squared. This is stated to be the correct measure (Guilford, J.P., Fundamental Statistics in Psychology and Education, 4th edition, 1965 p322) and other methods, such as comparing top and bottom thirds of the class, are short cut manual methods to approximate the same figures.) Mary Masson ---------------------- Dr Mary Masson Department of Chemistry University of Aberdeen Meston Walk, Aberdeen AB24 3UE Tel 01224 272931 Fax 01224 272921 email [log in to unmask] ```