Hashem Salarzadeh Jenatabadi and Ali Noudoostbeni control these factors [2]. End user plays as a vital rule achieving ERP success [3]. Knowledge [2] ...
Applied Mathematical Sciences, Vol. 8, 2014, no. 24, 1187 - 1192 HIKARI Ltd, www.m-hikari.com http://dx.doi.org/10.12988/ams.2014.4284
End-User Satisfaction in ERP System: Application of Logit Modeling Hashem Salarzadeh Jenatabadi Applied Statistics Department, Economics and Administration Faculty University of Malaya, Kuala Lumpur, Malaysia Ali Noudoostbeni Information Science Department, Computer Science Faculty University of Malaya, Kuala Lumpur, Malaysia Copyright © 2014 Hashem Salarzadeh Jenatabadi and Ali Noudoostbeni. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract The current study mainly aimed at validating the model of End User Computing Satisfaction (EUCS) in Enterprise Resource Planning (ERP) system in food production companies. In the EUCS model a Logit model on the seven items was applied. The tool used for the study was a questionnaire. A total of 126 questionnaires were collected from the ones given to ERP end users from food manufacturing in Malaysia, Taiwan, and China. Keywords: Logit model, Enterprise Resource Planning, End-User Satisfaction
1. Introduction ERP, in the last decade, has been used as a comprehensive system by many types of companies. It has been successful companies for implementing ERP and some others it as failed [1]. Therefore, so many studies have been done to find out the effective success and failure factors in ERP system and introduced techniques to
1188
Hashem Salarzadeh Jenatabadi and Ali Noudoostbeni
control these factors [2]. End user plays as a vital rule achieving ERP success [3]. Knowledge [2], development [4], acceptance [5], and satisfaction [6] are the most concerned studies in ERP end-user area. Szajna and Scamell [7] believes that end user satisfaction is one of the key factors leading to information system success and there is limited organized modeling to find out the effective factor on end user satisfaction in ERP system. This study explores the end-user satisfaction with an ERP, in the context of food manufacturing of China, Malaysia, and Taiwan, with the purpose to answer the following question: What are the antecedents of end-user satisfaction with a food manufacturing ERP? For this, the main objective of this work was to assess the end-user satisfaction regarding a strategic ERP system, which has been used for more than three years by those companies. Basic assumptions of normality together with continuous data, including dependent and or independent variables, are needed in the majority of the multivariate analysis techniques. Such a necessity also demonstrates itself in the collection of data and the measuring procedures of the research. Four-point Likert, ordinal and nominal scales are among the commonly used measurement scales; however, there are not considered suitable for multivariate techniques of analysis. Interval and ratio scales on the other hand provide better bases for a more thorough analysis. Considering the types and the frequency of applications of different multivariate analysis techniques, one comes to the conclusion that regression is among the highly frequently applied ones. Similarly in this situation, it appears that logistic regression is a greatly useful technique of analysis in order to model and discriminate problems in the satisfaction of the end user. Because of the alternative data distribution assumptions, it yields more accurate and appropriate results not only in model fit terms but also in analysis correctness. As a form of regression, a logistic regression model is one in which the outcome variable is dichotomous, or binary, and the independent ones are variables which are continuous, categorical or both.
2. Background of technical theory In order to predict a binary dependent variable from a set of independent ones, different multivariate statistical techniques, such as discriminant analysis or multiple regression, may be used. Yet, such techniques create problems once the dependent variables could have two values only. The assumptions required for the testing of the hypothesis in regression analysis would be violated if the dependent variable could have only two values. Standard lowest squares assumption of normal errors would be violated. Because E(y) shows a probability, it should range from 0 to 1. Nevertheless, it is not possible to interpret predicted values
End-user satisfaction in ERP system
1189
gained from multiple regression analysis as probabilities; therefore, they would not be limited to fall somewhere between 0 and 1 interval. Direct prediction of group membership is possible through linear discriminant analysis yet, for an optimal prediction rule there need to be an assumption of the independent variables’ multivariate normality together with equal variance covariance matrices in both groups. The model of logistic regression which we have utilized in our study would require much fewer assumptions than does the discriminant analysis model. Generally for categorical data, dichotomous data seem to be the most widely used. Binary logistic regression model makes it possible to include the numerous factors which influence dependent variables and that is the reason why it is so commonly applied to demonstrate in so many different areas. Density function of logistic regression has been provided as following formula: (
( )
) (
)
The probability of logistic regression function is the cumulative distribution function: ∫
)
((
( )
(
) )
Given by: (
) (
)
Where α=-µ/σ and β=1/σ. The binary logistic regression model can be written as: ( )
( ) ( )
( ) For above model β0 is the intercept; β1 is the parameter related to Xi1; β2 is the parameter related to Xi2; i=1,2,…,n; j=1,2,…,k; Yi=1,0. For a binary response Y and the explanatory variable X, let πi represent the probability of “success.” To simplify the notation when using a logistical distribution, in this study we use πi (x) to represent the conditional mean of Y given x: ( )
(
)
1190
Hashem Salarzadeh Jenatabadi and Ali Noudoostbeni
( )
(
)
Such that: (
(
)
)
The transformation of π(xi), which is central to this study of logistic regression, is the logit transformation. This transformation is defined as: (
( ) ) ( )
(
( ) ) ( )
(
)
Where Zi is the logit of the probability of “success” for the i-th level of X.
3. Methods and Discussion A quantitative research survey is employed to examine the hypotheses proposed in the research framework. The data collection period spanned between October 2012 and February 2013 for a period of five months. The prepared questionnaires were distributed among 387 End users of ERP selected from food manufacturing in Malaysia, Taiwan, and China. Only 126 completed questionnaires returned (without missing data) which provided this study with a response rate of 33%. The research model includes one dependent variable (End-user satisfaction; 1=satisfy 0=not satisfy) and seven independent variables contain GENDER [1=man, 0=woman]; MARITAL [0=single, 1=married]; EDJUCATION [1=under diploma,2=diploma, 3=bachelor, 4=master and above]; EXPERIENCE [working experience] ; INCOME [average income per month]; COMPUTER [ computer familiarity from 1-4; Very little, Little, Some, and A lot]. (
( ) ) ( )
(
( ) ) ( )
End-user satisfaction in ERP system
1191
Table 1: Results of Logit model Variable Gender Age Education Level Martial experience income computer Constant
B -.341 .328 .425 1.549 .147 .687 .781 -6.747
S.E. .520 .229 .230 .514 .335 .296 .233 1.472
Wald .430 2.053 3.417 9.075 .194 5.391 11.238 20.999
Sig. .512 .152 .065 .003 .660 .020 .001 .000
Exp(B) .711 1.388 1.529 4.707 1.159 1.987 2.184 .001
In recent years, large and small organizations have born huge costs by investing hugely on the installation and launching ERP systems; such investments in some companies have been over millions of dollars. The managers of such companies have been completely aware of the effectiveness of this system for their firm and believe that by launching this system they could increase their data potential and the quality of their required real data and based on that they can apply a reengineering in order to improve the performance of the internal departments of their company. However, one of the existing problems of such companies in using this system is that the end-users are not really willing to use this system. This research aims at investigating the effective factors on the end-user’s satisfaction in food production industry. The results can identify the factors which influence the end-users’ satisfaction with the ERP system and by applying effective programs on the change of effective factors, it is possible to increase the satisfaction in the end-users. Based on table 1, it can be seen that the variables computer, marital status and income have a positive and meaningful effect on the end-user satisfaction. Therefore, with an efficient program, adapting these three variables suitably the level of desired satisfaction could be obtained. For example, it is possible to choose those people as end users who are married and possess a lot of computer familiarity. Even by changes in such people’s wages, one could increase their satisfaction.
References [1]
Noudoostbeni, A., N.M. Yasin, and H.S. Jenatabadi. To investigate the success and failure factors of ERP implementation within Malaysian small and medium enterprises. in Information Management and Engineering, 2009. ICIME'09. International Conference on. 2009. IEEE.
1192
[2]
[3]
[4]
[5]
[6]
[7]
Hashem Salarzadeh Jenatabadi and Ali Noudoostbeni
Noudoostbeni, A., et al., An effective end-user knowledge concern training method in enterprise resource planning (ERP) based on critical factors (CFs) in Malaysian SMEs. International Journal of Business and Management, 2010. 5(7): p. P63. Noudoostbeni, A., N.M. Yasin, and H.S. Jenatabadi. A mixed method for training ERP systems based on knowledge sharing in Malaysian Small and Medium Enterprise (SMEs). in Information Management and Engineering, 2009. ICIME'09. International Conference on. 2009. IEEE. Spahn, M., S. Scheidl, and T. Stoitsev, End-User Development Techniques for Enterprise Resource Planning Software Systems. End-User Software Engineering, 2007. 7081. Youngberg, E., D. Olsen, and K. Hauser, Determinants of professionally autonomous end user acceptance in an enterprise resource planning system environment. International Journal of Information Management, 2009. 29(2): p. 138-144. Calisir, F. and F. Calisir, The relation of interface usability characteristics, perceived usefulness, and perceived ease of use to end-user satisfaction with enterprise resource planning (ERP) systems. Computers in Human Behavior, 2004. 20(4): p. 505-515. Szajna, B. and R.W. Scamell, The effects of information system user expectations on their performance and perceptions. Mis Quarterly, 1993: p. 493-516.
Received: February 5, 2014
