Validity, Reliability, and Normality Test Data Analysis

40

3.7 Validity, Reliability, and Normality Test

According to Ghozali, 2001 validity test is a tool which is used to measure validation of questioner. Questioner is valid if the range is more than 0.30 question of questioner can describe something that will be measured by questioner. Validity test used SPSS program. In order to validate a measurement instrument, it must first be subjected to test of both validity and reliability. Since most of the questions in the questionnaire were self administered, it is felt necessary to examine the validity of the measure. One of the important steps in data analysis is to confirm whether or not the variables representing responses to questions are uniquely associated to the theoretical dimension of the variables of interest Sekaran, 2000. According to Ghozali 2001:41 is measuring instrument to measure a questioner which represent indicator of construct variable. The most popular test of inter item consistency reliability is the Cronbach’s coefficient alpha Cronbach’s alpha; Cronbach, 1946; cited from Sekaran 2003. The way to calculating of a data reliability level is using Cronbach alpha is between 0.60 – 1.00. Normality test can be used the Kolmogorov Smirnov test, whereby if the sign value 0.05, then it can be concluded that the data variables were tested with the normally distributed. 41

3.8 Data Analysis

This study is intended to test a model that explained the effect of justice perception on job satisfaction and its impact for nurse’s motivation to answer the hypothesis, the data will be analysed using statistical package for the social sciences SPSS 16.0 and for structural equation model SEM. SPSS is needed to analyse the respondent characteristic in represent the frequency and percentage of respondent data. Beside that this tools also used to determine the validity, reliability, and in this research normality measure using AMOS. The data will be analysed using structural equation model SEM by AMOS as software application. This software provides information about goodness-of-fit model and relationship among the hypothesis. Moreover SEM was commonly used measures of fit include:  Chi-Square a fundamental measure of fit used in the calculation of many other fit measures. Conceptually it is a function of the sample size and the difference between the observed covariance matrix and the model covariance matrix.  Akaike information criterion AIC o A test of relative model fit: The preferred model is the one with the lowest AIC value. o o where k is the number of parameters in the statistical model, and L is the maximized value of the likelihood of the model.  Root Mean Square Error of Approximation RMSEA 42 o Another test of model fit, good models are considered to have a RMSEA of .05 or less. Models whose RMSEA is .1 or more have a poor fit.  Standardized Root Mean Residual SRMR o The SRMR is a popular absolute fit indicator. A good model should have an SRMR smaller than .05.  Comparative Fit Index CFI o In examining baseline comparisons, the CFI depends in large part on the average size of the correlations in the data. If the average correlation between variables is not high, then the CFI will not be very high 43 Table 3.2 Evaluation of SEM with Goodness of fit Measure Types of Measure Goodness of fit Measures Recommended Level of acceptable Fit Absolute Fit Measure Goodness of fit index GFI Root mean square error of approximation RMSEA Greater than .90 Under .08 Incremental Fit Measure Adjusted goodness if fit index AGFI Turker – Lewis index TLI Normed fit index NFI Comparative Fit Index Greater than .90 Greater than .90 Greater than .90 Greater than .90 Parsimonious Fit Measure Normed chi- square χ2df AIC Lower limit 1.0 Upper limit 2.03 3.0 or 5.0 Smaller positive value indicate parsimony Source: Tabachnick and Fidell 2000; Hail et al. 1998; Byrne 2000. 44

CHAPTER IV ANALYSIS AND RESULTS

4.1 Survey Results

This chapter will discuss analysis of research result related with job stress, organizational support, job satisfaction and performance of nurses who work at private hospital in Padang. The result of this survey can be seen from the number of questioner have been distributed to nurses who work at some private hospitals in Padang. Data was distributed to nurse through their head of installation and chief of room for each installation in private hospital, and take it back about 2 weeks to fill in the questioners. This is the table 4.1 that show about the number of questioner. Table 4.1 Survey Result Survey Number of questionairs Distributed 164 Returned 150 Not Returned 14 Analysed 150 Source: Processed from questionnaire by using SPSS From the table 4.1, 164 questionnaires was distributed to nurses who work at some private hospital in Padang. Others 14 questionnaires were not returned back to researcher because of some nurses were not in Padang, absenteeism and vacation, and some nurses unfilled it because they did not have