Measurement of Organizational Citizenship Behaviour; Reliability and Validity in Sri Lankan Context

This paper investigates the goodness of the measure of organizational citizenship behaviour in terms of reliability and validity. Organizational Citizenship Behaviour considers as the extra role or voluntary behaviour not directly or explicitly recognized by the formal reward system, and that in the aggregate it promotes the effective functioning of the organization. However, more research studies on OCB focus on western context while similar effort in Sri Lankan context rather scant. Data were gathered through the survey by distributing structured questionnaire from public sector organizations. As per the discussion basically, two main criteria called reliability and validity have to be achieved to confirm the goodness of the measure. Internal reliability and composite reliability scales were commonly employed to asses construct reliability of the intended constructs. However, convergent validity achieved through Average Variance Extracted (AVE) and factor loadings. Discriminant validity can be evaluated by assessing the cross loadings among constructs, Fornel-Larcker criterion, and HeterotraitMonotrait Ratio of correlation (HTMT). According to the derived outcomes implications regarding the goodness of measure were discussed and revisions of measurement in Sri Lankan context were presented.


Introduction
Much of the scholarly concentration in OCBs stems from the extensive belief that OCBs improve the efficiency and effectiveness of organizations (Organ, Podsakoff & Mackenzie, 2006).Organizational Citizenship Behaviour defines as individual behaviour that is discretionary, not directly or explicitly recognized by the formal reward system, and that in the aggregate promotes the effective functioning of the organization (Organ, 1988).In the first part of the original definition, OCB treats as the notion of discretion, which means that engagement in OCB is completely voluntary (Organ et al., 2006) or not an enforceable by supervisors or superiors, it is rather a matter of persons choice, such that its omission is not generally understood as punishable (Organ ,1998).Further, it highlights that task performance and OCB are separated and OCB is not explicitly recognized by the formal reward system (Organ, 1997).
OCBs include both behaviours that are directed toward specific persons (e.g.interpersonal helping), as well as more impersonal forms of conscientiousness and workplace involvement that contribute to organizational effectiveness (e.g.sportsmanship, organizational loyalty, organizational compliance, individual initiative (Organ et al., 2006)).These extra-role behaviours are likely to facilitate social and psychological contexts that support task performance in the organization (Organ, 1997).As a whole, OCB could contribute to enhancing employees and managerial productivity, attract retention, improve coordination, reduce friction, viability of performance and increase adaptation to environmental changes (Podsakoff, MacKenzie, Paine, & Bachrach, 2000).

Objective of the Study
To find out the internal reliability and the validity of the concept of organizational citizenship behaviour.Katz and Kahn's (1966) "The social psychology of organizations" pave the path to thinking about OCB within the context of open system as a model of organization theory.Katz (1964) explained that behaviours which are helpful, innovative and cooperative are essential for organizational operations.et al., 2005).Courtesy has been defined as voluntary behaviours that aim at preventing work related conflicts with others (Law et al., 2005).Specifically, it attempts to mitigate interpersonal problems from occurring by giving advance notice and try to prevent from encountering unpleasant surprises.Finally, Sportsmanship has been defined as a person's desire not to complain unnecessarily and being positive and tolerant when experiencing the inevitable inconveniences.On some occasions they are willing to sacrifice their own personal interest for the sake of group interest.These behaviours reflect an employee recognition of being a part of the organization (Podsakoff et al., 2000).

Measurements
The first decision to be made in operationalizing a construct is to decide on what is the intended level of measurement.Different measures have been designed and developed by researchers.The first scale was designed by Bateman and Organ (1983) which was used to measure Organizational citizenship behaviour.According to them, "OCB include any of those gestures, that lubricate the social machinery of the organization but that do not directly in here in the usual notion of task performance."This scale was a 30 item OCB scale that comprises of variety of types of OCB like cooperation, altruism, compliance, punctuality, housecleaning, protecting company property, conscientiously following company rules and dependability.A 16item scale was developed by Smith, Organ and Near (1983).According to them, "Citizenship Behaviours comprise a dimension if individual and group functioning.The latter was regarded as a function of the formal organization and the logic of facts."This scale comprises of two factors namely Altruism and Generalized Compliance.The third scale was given by Podsakoff, Mackenzie, Moorman, and fetter (1990)

Assessment of Construct Reliability
PLS-SEM or partial least squares path modeling is a variance-based structural equation that has become very popular in recent years (Henseler, Hubona &Ray, 2016).In PLS analysis, the first step is to assess the measurement model or the outer model.The two main criteria used in PLS analysis to assess the measurement model or what is alternatively called the outer model include validity and reliability (Ramayah, Lee & In, 2011).The individual item reliability, construct internal consistency and construct validity are considered in assessing the outer model in PLS.Reliability is a quality criterion of a construct; it requires a high level of correlation among the indicators of a particular construct (Kline, 2011).The internal reliability of a construct is said to be achieved when the Cronbach's Alpha value is 0.7 or higher (Nunnally & Beinstein, 1994).However, Composite reliability more concern on individual reliability referring to different outer loadings of the indicator variables (Hair, Hult, Ringle, & Sarstedt, 2017).The cut off for composite reliability is the same as any measure of reliability and score between 0 .6 and 0.7 is a good indicator of construct reliability (Hensele &, Sarstedt, 2013).

Assessment of Validity
Validity concerns the soundness of the accuracy of a measure or the extent to which a score truthfully represents a concept (Zikmand, Babin, Carr &Griffin, 2013).In order to achieve validity analysis, two kinds of validity tests were performed in PLS on the measurement scales namely: convergent validity and discriminant validity (Sekaran & Bougie, 2010).Convergent validity is the extent to which a measure correlates positively with an alternative measure of the same construct.In examining the convergent validity of a measure in PLS, the average variance extracted (AVE) and item loadings are assessed (Hair et al., 2013).Discriminant validity is concerned about the uniqueness of a construct, whether the phenomenon captured by a construct is unique and not represented by the other constructs in the model (Hair, Ringle, & Sarstedt, 2013).Discriminant validity can be evaluated by assessing the cross loadings among constructs, by using Fornel-Larcker criterion and Heterotrait-Monotrait Ratio of correlation (HTMT).

Methodology
The study involves a testing of measurements and it is quantitative, deductive in nature.Employed survey method and cross sectional study in type with minimum researcher interference.Unit of analysis is the individual employees belonging to the front line worker employees of the public sector organizations in Sri Lanka.The employees' OCB is measured by using 24 item OCB, which was developed by the Podsakoff and colleagues (2000) based on the Organ's five-dimensional taxonomy.The five-point Likert scale designed from 1 (strongly disagree) to 5 (strongly agree).The purpose of the study is to test the reliability and validity of measure of organizational citizenship behaviour in Sri Lankan context.A sample of 296 employees from the public sector organizations was drawn with disproportionate stratified random sampling method.Questionnaires were distributed to the respondents.This study recorded a nearly 67% of response rate.It seems that the external validity of the finding of this study can be held high with this higher response rate.

Discussion
The construct reliability and validity test were conducted to evaluate the measurement of the construct of OCB.Composite reliability as a greater measurement instrument will explain the internal consistency of the manifest variables concerning items differently.The table 1.4 illustrated that all loadings are above the cutoff values and composite reliability of every construct was well above ranging from 0.788 to 0.809 as per the suggested 0.70 threshold.Further, Convergent validity examined concerning the item reliability or the factor loadings and Average Variance Extraction of the constructs.AVE value equal or higher than 0.50 indicates that on the average, the construct explained more than half of the variance of its indicators.As such, the rule of thumb is that an AVE value greater or equal to 0.50 is acceptable (Hair et al., 2013;Barclay, Higgins, & Thompson, 1995).However, out of the initial 24 items eight indicators omission from the construct of OCB as OCBA1 (0.608), OCBA2 (0.577) ,OCBS4 (0.645) ,OCBS5 (0.616) ,OCBCON1 (0.649) ,OCBCON5 (0.618) ,OCBCUR4 (0.637),OCBCUR (0.636) to increase the composite reliability and the AVE.In the present study, the values of AVE of all revised variables exceeded the threshold which indicates the good construct validity of measures.In order to confirm construct validity of the outer model, the discriminant validity is indispensable.Discriminant validity explains about the extent to which a construct is truly distinct from other constructs by empirical stands.At first, in order to achieve discriminant validity, the loadings of the construct must be high on itself and low on other constructs (Vinzi, Henseler, Chin & Wang, 2010).
The following table (Table 02) clearly explain that all the loadings under the construct itself are well above than the cross loadings.(Fornell & Larcker, 1981).The square roots of AVE coefficients are presented in the correlation matrix along the diagonal.The squared root of each constructs' AVE should be greater than its highest correlation with any other construct to evidence discriminant validity (Hair et al., 2013).

Findings and Conclusion
This study is intended to test the goodness of measure for organizational citizenship behaviour construct in the Sri Lankan context.The purpose of testing the measurement of organizational citizenship behaviour construct will add to the contextual validity in a different context.The outcome of the reliability test shows that measurement has a better internal consistency in measuring the construct and it is stabilized.Further, in line with construct validity; convergent and discriminant validity added more confirmation on the accuracy of the measurement.Convergent validity tested the degree to which indicators of a particular construct converge or share a high proportion of variance in common (Hair et al., 2010) while Discriminant validity is referred to the uniqueness of a construct, whether the phenomenon captured by a construct is unique and not represented by the other constructs in the model.The original item structure has 24 items to measure and that was not totally consistent with the pre-test model of the present study given the fact that several items had to be dropped due to low loading .Therefore, the finding of this study exhibits the differences on the dimensionality of OCB in a different context from where it was mostly tested.
Any way the revised model results clearly shows that all manifest variables measure the same construct achieving convergent validity and exhibits the discriminant validity and are distinct from one another.It can be an impetus for increase studies in the Sri Lankan context given the fact that more research required for its conceptual clarity and validity.