Evaluation of Hajj Instrument ( HAJI ) Psychometric Properties Using Rasch Measurement

Hajj Instrument (HAJI) was developed to determine hajj pilgrim’s wellness. This study used Rasch measurement to evaluate the psychometric properties including validity and reliability of the HAJI. The respondents involved in this study were 300 comprised of Malaysian hajj pilgrims. HAJI consists of eight constructs namely physical care, physical activity, healthy eating, knowledge, mental toughness, intrapersonal, interpersonal and relationship with Creator and natures. Validity of each construct and content was determined through dimensionality, item fit and item polarity while the reliability was achieved by administered person and item separation. The results showed that the reliability for both item and person were 0.99 and 0.96 respectively. Besides, there were no items need to be dropped based on PTMEA CORR and INFIT MNSQ results. The study revealed that the items of HAJI fit the Rasch model as well as able to measure hajj pilgrim’s wellness.


Introduction
There are extensive literatures on the definition of wellness but few researches about the wellness assessment (Anspaugh, Hamrick, & Rosato, 2008).Wellness assessment is a tool to assist human in establishing positive lifestyle behaviors, execute early health interventions or reduce other health risks (Haddad, Owies, & Mansour, 2009).Several researchers have discussed on the difficulty of assessing the dynamic nature of wellness as well as insufficiently of the existing measurement tools (Rachele et al., 2013;Renger et al., 2000).There are several methods have been developed to measure wellness such as Life Assessment Questionnaire (LAQ) (Tearnan & Ross, 2012), TestWell (Brown, Applegate, & Yildiz, 2015), Perceived Wellness Survey (Rothmann & Ekkerd, 2007), Optimal Living Profile (von Guenthner & Hammermeister, 2007) and Wellness Inventory (Roscoe, 2009).However, those methods assess wellness in general which means not specific to certain events and age.Therefore, the results will not fully representative the general population and may not accurately address pilgrims' wellness issues or needs at all if use those assessments.Nevertheless, there is argument that some of these assessments have either good reliability or validity (Brent & Carlson, 2014).Thus, study aims to evaluate HAJI reliability and validity in order to be used as a measurement instrument for hajj ritual.

Method
This study used questionnaire and were distributed to 300 respondents consist of hajj pilgrims.The respondents were randomly chosen from six mosques which organized Hajj courses in Johor Bahru district.The questionnaire consist of 72 questions that were divided into eight construct namely physical care, physical activity, healthy eating, knowledge, mental toughness, intrapersonal, interpersonal and relationship with Creator and natures.Rasch measurement (RM) was used to analyze the validity and reliability of HAJI based on psychometric standard criteria including item dimensionality, item polarity and item fit analysis.Instrument calibration scale also was administered to assess the suitability of the scale used in HAJI.Four Likert scale was used in HAJI consist of Strongly Agree (4), Agree (3), Disagree (2) and Strongly Disagree (1). www.ccsen

Finding
Validity a followed.

Dimension dimension unexplaine dimension
It can be unexplaine dimension  Results show that value of item reliability was 0.99 whereas item separation value is more than two (10.04) as shown in Figure 2. While, Figure 3 shows person reliability was 0.96 whereas for the person separation the value was five.Thus, it can be conclude that HAJI has strong and acceptable reliability (≥ 0.8) as suggested by Aziz et al., (2008), Bond & Fox (2015) and Mamat, Maidin, & Mokhtar (2014).This means respondents involved represent actual characters that need to be tested.Besides, result for separation was good and it shows variety of skills in answering HAJI (Smith, 2000).

Item Polarity Analysis
Item polarity is necessary in measuring the constructs validity.It is similar to factor analysis function where it is used to access the relationship of the items in measuring the construct.The criteria of good correlation are the values PTMEA should be > 0.20 (Bond & Fox, 2015).Table 1 shows there were no value of negative correlation and all PTMEA of each items is > 0.20.This indicates that there are no mistakes in data entry or miscoded items.Table 2 shows the summary of item polarity analysis.

Item Fit Analysis
In order to analyze appropriateness of HAJI items, each item were analyzed separately.Each measured item shows the information of mean square (MNSQ) which can used to identify misfit item.According to Bond & Fox (2015), acceptance value of MNSQ for infit analysis should be 0.4 < x <1.5 and ZSTD values range between -2 and 2. Table 1 shows analysis of item fit for all items.From the table, it can be seen that infit MNSQ values for all items was within the standard range of RM.It means all the 72 items fit the construct and may not be drop.Table 3 shows the summary of item fit analysis.
Figure 1.Ana Figure 3 and ersection point (6119), 2 (197 start from neg th of Rasch m n this study, th ion for three a by respondent than 1.4 and le

Table 1 .
Analysis of item polarity and item Aziz et al.,lysis could validate the scale used by made zero calibration setting.Rasch analysis determines validity of respond possibility to spread fairly between specified scales(Alagumalai, Curtis, & Hungi, 2005;Aziz et al.,