1. Introduction to Rasch Measurement Model Intangible Measurement 231804-P Mohd Saidfudin Masodi Measurement and Psychometry Integrated Advance Planning Sdn. Bhd. [email_address] +60 1240 2821

2. The Speaker: An introduction… MOHD SAIDFUDIN MASODI, B. Arch. (Aust.); IRCA (Lon.) ISO QMS Lead Assessor, Measurement / Psychometry Consultant Trained by Nigel Bauer Ltd., UK and Rheinisch-Westfalischer Technischer Uberwachungsverein A.G. International (RWTUV, Germany) and AOTS Japan with wide experience in conducting assessment in local and multi-national corporate organisations specifically the institutions of higher learning. Expertise: Rasch measurement and probabilistic statistical methods, Quality Audit. Research Area: Measurement, evaluation and Statistical Analysis in (MESA) Learning Outcomes (LO) and learning culture in institution of higher learning using Bloom’s & SOLO Taxonomy. Publications: Co-published 3 book chapters, eight monographs, and over 30 international conference proceedings and articles in prime international refereed journals i.e.IEEExplore, WSEAS Transactions, NAUN etc. Special Appointments: Member in the Board of Study for the programs International Labour Organisation, Executive Management Program, Master in Protective Security Management, IIUM; GiatMARA and Human Resource Dept, State of Sabah . Other credentials: Best Paper Award in ICEED 2009 (Eng’g Education) Best Visiting Lecturer UTM_SPACE 2007, 2008. Invited as plenary speaker in several international conferences and many faculty talks. Conduct training and short courses regularly; in area of Strategic Planning using Blue Ocean Strategy and Balanced Score Card, Testing & Evaluation and KPI QMS Internal Audit and QA/QC in the manufacturing industry.

3. 01:25 AM M O D U L E - 2 I N T R O D U C T I O N TO RASCH UNIDIMENSIONAL MEASUREMENT MODEL Integrated Advance Planning Sdn. Bhd. [email_address] +60 1240 2821

11. 01:25 AM A valid measurement must meet five (5) criteria: Measurement of Psychological Construct 1. Linear Scale: AMOUNT 2. Reasonable numerical values: ACCURACY 3. Empirical Coherence: Response, Item and Construct VALIDITY 4. Incorporate parameter separation: REPLICABLE 5. Overcome missing data: PREDICTIVE (Wright & Mok, 2004) Mohd Saidfudin Masodi +60122402821 [email_address]

13. 01:25 AM Concept: Liken Person’s Ability and Height of High Jump Bar Person’s Ability Varying height The higher, the more difficult Developing a Measurement Construct In any measurement endeavour, the aim is to put enough stepping stones along the path to represent all the stepping points useful for the measurement purposes; from little to much development. (Bond & Fox, 2001) Georg Rasch (1901-1980) B -DIFFICULT A -EASY Mohd Saidfudin Masodi +60122402821 [email_address]

15. 01:25 AM e.g. On a graduation day, what is the likelihood of a lady liking a piece of rose as your giving ? Perhaps 30:70 Compare if you send a bouquet instead. It increases to 60:40; and so forth if you put a Fererro Roche.. the chances gets better. 10 90 In Rasch Model, a turn of event is seen as a chance; a likelihood of happenings hence a ratio. Rasch Unidimensional Ruler 10 -2 -2 30 70 60 40 50 50 99 1 1 99 10 0 10 2 0 2 -1 1 indices logit Now, we already have an instrument with a unit termed ‘ logit ’. Mohd Saidfudin Masodi +60122402821 [email_address]

16. 01:25 AM Student.1 1. But, atypical test result tabulation only rank the students from the highest score in descending order 2. Need to assess beyond raw score. Rasch sorts further according to response pattern in descending order; called ‘Guttman scale’. Rasch Model Measurement Theorem Student.7 S-03 S-05 Q1 Q15 Q16 Q30 Q31 Q50 10111 0 11111111111 11111111111111111 111111111111001 = 48 1010010001111111 11111111111111111 111111111011111 = 43 10111111111111111 1110111111100100 01101010001101 = 33 10111111111011111 1111111111010100 10110100000011 = 33 1011 0 111111111111 1 0 111111 0 1 00 111 0 0 00 1 0000 1 0000 1 = 33 10111111111111111 111101100100010 01000000000001 = 27 10111111111111101 110101000100010 0000000000 1 001 = 24 Mohd Saidfudin Masodi +60122402821 [email_address]

17. SMART POOR T1. Persons who are more able / more developed have a greater likelihood of correctly answer all the items / able to complete a given task. 7 6 4 3 0 Rasch Model Measurement Theorem TOTAL RESPONSE S-01 S-07 S-05 S-03 5 S-02 EASY ITEMS Q3 q6 Q1 Q7 Q5 DIFFICULT ITEMS Q4 Q2 11111 0 11111111111 11111111111111111 11111111111100 = 48 11 0000 11111111111 11111111111111111 11111 11111110 0 = 43 111111111111111 1 1111111111110011 11100010000000 = 33 111110111111111111 0111111111101001 00110100000000 = 33 11111 0 1 00 111 0 11 0 1 1 0 111111111111 0 1 0 0 11 0 1 00 1 0 1 0 1 0 = 33 11111111111111110 0111011101000100 010000 0000000 = 27 11111111111111101 1101110100100100 00000000000 1 00 = 24 T2. Easier items / task are more likely to be answered correctly by all persons. Mohd Saidfudin Masodi +60122402821 [email_address]

18. 01:25 AM Rasch Measurement Model Theorem Two (2) propositions appears: 1. Persons who are more able / more developed have a greater likelihood of correctly answer all the items / able to complete a given task. 2. Easier items / task are more likely to be answered correctly by all persons. In summary; the additive correlation is : Person Ability Pr (Success ) Difficulty of a given task = - Mohd Saidfudin Masodi +60122402821 [email_address]