Interpersonal Power Inventory (IPI) as well as examine and ... assessment tool to diagnose and guide leader .... appeal, consultation, negotiation, exchange,.
International Conference on ArtS and Sciences – 2014 (IJAC’14) EXPLORATORY FACTOR ANALYSIS: CONCEPTUALIZATION, RELIABILITY AND VALIDITY OF INTERPERSONAL POWER INVENTORY Abdul Aziz Rozilah Universiti Teknologi MARAz, Malaysia Hassan Narehan, Shereen Noranee, Noor Azlina Mohd Yunus, Siti Nurul Akma Ahmad Universiti Teknologi MARA, Malaysia Abstract—The aim of this study is to adapt and
analyze the validity and reliability of Interpersonal Power Inventory (IPI) as well as examine and develop the psychometric properties of the IPI construct which integrates 5 subscales – reward, coercive, expert, referent and legitimate, information powers with 3 of these further differentiated; reward (personal, impersonal), coercion (personal, impersonal), legitimate (position, reciprocity, equity, dependence) to be used in the Malaysian organizational context. A questionnaire survey was administered to 101 executives working directly under Human Resource Managers in the state-owned organizations, known as Government Link Companies (GLCs). A factor analysis was initially conducted on 44 items of IPI to explore the structure underlying the set of questions designed. The results of exploratory factor analysis (EFA) showed and confirmed a fivefactor structure: Reward, Coercive, Legitimate, Expert, and Referent leaving only 25 items from the original 44 items. EFA results also confirmed that the measurement scale used in this study satisfactorily met the standards of validity and reliability analysis and showed acceptable internal consistency reliability for the overall and the five specific subscales of power factor. The IPI construct provides a multi-dimensional assessment tool to diagnose and guide leader social power in the Malaysian organizational context. Keywords-Reward Power, Coercive Power, Legitimate,Expert Power, Referent Power
I.
INTRODUCTION
For many decades the existence of power is a crucial topic in understanding human behavior and has a significant impact on the implementation of the strategies, policies, and decision making in organization. As power has been studied from varies perspective and described by many different sociological and psychological theories, its intricacy and understandable provide a wide variety of definitions and operationalizations [1]. A commonly cited definition is that of [2], [3] and [4] who define social power as the ability on one person to change or control the behavior, attitudes, opinions, objectives, needs, and values of other person. This definition implies that research on power is limited to the influence of one individual (leader) over another individual (follower). Several classifications of leader or supervisory power have been set forth however, [5], [3] and [6] conceptualized that social power is the resources of one person have to influence another person to do what that person would not have done otherwise. Similarly [6] made a valuable contribution to power conceptualization and defined power as the capacity to influence individuals to do something they would not have done has they not been influenced. Several other investigators who have used different methodology approaches to operationalize power argued that the number of basic base of power may be greater than the five original formulation [5]. To answer some of these concerns [3] developed social power measures and presented an interpersonal power Inventory (IPI). In IPI, while maintaining the basic conception of six bases of power, evidence was cited from theory and research to point to the need to make
International Conference on ArtS and Sciences – 2014 (IJAC’14)
differentiations among some of the bases of power. IPI constructs integrate 11 subscales – maintaining the basic conception of six bases of power: reward, coercive, expert, referent, information and legitimate powers with 3 of these further differentiated; reward (personal, impersonal), coercion (personal, impersonal), legitimate (position, reciprocity, equity, dependence). Prior to the development of an interest in social power and examination of these approaches, there is dearth study on IPI that can be used in the Malaysian organizational context.
II. LITERATURE REVIEW Over the years, the study of power sources has become a central concept in organizational psychology, as well as in other analyses of social interaction. Evidences showed numbers of instruments developed by other investigators were derived from a priori proposition made by [7] typology who suggested 5 bases of power taxanomy: (a) reward power – promise of monetary or nonmonetary compensation; (b) coercive power – threat of punishment; (c) legitimate power – drawing on one’s right to influence; (d) expert power – relying on one’s superior knowledge; and (e) referent power – based on target’s identification with influencing agent [8]. These conceptualizations derived from the early work of [9], who posited a model of human behavior based upon forces within one’s life space. [10] has reconceptualized social power to the extent that the supervisor’s control over subordinate’s behavior depends on the magnitude of the force which supervisor can bring on subordinates and upon subordinate’s resistance. In a later elaboration, based on presentation of persuasive material or logic, informational power was also considered as a basis of power as it conceived as distinct from expert power. However, despite the broad acceptance of this typology of power in survey texts of social and organizational psychology, [8] identified some serious methodological and substantive concerns. Concerns have been raised in regard to conceptual overlaps between categories, inconsistent operational definitions,
and, at times, low agreement between observation and prediction [3]. For example, allusions to expertise may involve many elements of referent and informational power. This problem was heightened by the fact that differing power bases are often used in combination (e.g expert and informational power), that influencing supervisors tend to reveal profiles of powers using unique constellations of bases rather than simple clear preferences for individual bases [11], [12], and [13]. Finally, several other investigators using different methodological approaches argued that the number of power sources may be greater than the five or six in the original formulation [5]. [14] identified another two power sources; persuasiveness and charisma, which did not appear in the original French and Raven Taxanomy. However, their definition of persuasiveness overlaps with Raven’s [15] definition of information, and charisma was seen as a projected character trait which was closely related to referent power. [16] reported a ninefactor taxanomy (rational persuasion, inspirational appeal, consultation, negotiation, exchange, personal appeal, collision, legitimating, and pressure). In IPI model which derived from the work of , [14], [17], and [18]while maintaining the basic conception of six bases of power, evidence was cited from theory and research affirmed the need to make differentiations among some of the bases of power . Coercion and reward power which were conceptualized in terms of “impersonal” threats or promises of reward are now distinguished from personal coercion and personal reward [3]. Raven’s new approach further differentiated coercion, reward and legitimate power (coercion power, personal coercion power, reward power, personal reward power and added (a) legitimate position power – supervisor’s right to prescribe behavior for the subordinates; (b) legitimate reciprocity – stems from the target’s obligation to comply with supervisor’s request after the supervisor has done something positive for the subordinate; (c)
International Conference on ArtS and Sciences – 2014 (IJAC’14)
legitimate equity – draws on the equity norm where supervisor demands compliance to compensate; and (d) legitimate – dependence – based on a social responsibility norm which obliges one person to assist another who is in need of assistance ). IPI model, [3] presented a situation where a subordinate, though initially reluctant, eventually complies with the supervisor’s request. As a result, IPI is developed with 11 bases of power: the original 6 French and Raven [7] bases of power, with 3 of these further differentiated; reward (personal, impersonal), coercion (personal, impersonal), legitimate (position, reciprocity, equity, dependence), expert, referent, and information. It was reported that each of these power bases contributed to a supervisor successfully influencing a subordinate in a series of hypothetical situations. The internal consistency of the items which made up the 11 power bases proved adequate. In most study, social power is mainly measured by French and Raven Power Taxonomy because there is extensive evidence of psychometric properties of the instruments in various different samples. Nevertheless, [3] argued that a number of scales developed by others to measure the original 5 or 6 bases of power often failed to fully satisfy the conceptual definitions of the bases of power as originally presented. For that reason [3] developed a new instrument of social power called IPI with eleven 11 bases of taxanomy. However, very few attempted to review and assess the IPI conceptualization of social power bases in Malaysia. Considering the importance of the social power construct, the intricacy and understandable in its definition and quantification, it warrants an intensive research on the IPI construct that is suitable and applicable to the Malaysian organizational context. Therefore, in accessing the construct validity of IPI an exploratory factor analyses (EFA) was utilized. An EFA was conducted to test the validity and reliability of IPI as well as identifying obvious
poor loadings of items under each dimensio [19]. Overall, the EFA purpose of this study was conducted to examine the stability of the factor structure and provide information that would facilitate the refinement of a new IPI measure. Subsequently, a scale development which includes an assessment of the psychometric properties of the scale was constructed. It is necessary to administer the potential items to a representative sample in order to examine how well the items confirm expectations related to the structure of the measure in question [19]. The scale has been consistently administered and the psychometric properties of this scale have been highly reliable.
III. MATERIALS AND METHODS A. Procedures
The IPI Instrument was adapted from [3]. Some of the items in IPI were changed for the purpose of the future study in the Malaysian organizational context. Researchers decided to change the word (a) “My Supervisor” in IPI scale to ‘My Manager;’ (b) Under the legitimacyposition, one item has been changed from “After all, he/she was my supervisor” to “As she/he is my manager, I feel some obligation to go along with her/him.” In expert power construct, all items has been reword/paraphrase based on the theory and literature review: (a) “My supervisor probably knew the best way to do the job” to “My manager handles any work situations proficiently and skillfully” (b) My supervisor probably knew more about the job than I did” to “My manager is the best person to handle any situations in this organization” (c) I trusted my supervisor to give me the best direction on this” to “I trust my manager to give me the best directions pertaining to work” (d) My supervisor probably had more technical knowledge about this than I did” to “I feel my manager has more technical knowledge pertaining to work in this organization”; (d) one item has also been added to Expert Power “My manager is given an opportunity to make any decisions based on his/her knowledge and
International Conference on ArtS and Sciences – 2014 (IJAC’14)
expertise”. In the coercive power constructs, one item has been reworded from “My supervisor may have been cold and distant if I did not do as requested” to “My manager gives silent treatment to me if I do not comply with him/her”. Finally, in the referent power constructs, one item has been added into the dimension of Referent power (i.e “I admire my manager”) based on the previous theory and literature review. Two items have been changed: (a) “I respected my supervisor and thought highly of him/her and did not wish to disagree” to “I respect my manager”; and (b) “We were both part of the same work group and should have seen eye-to-eye on things” to “Sometimes I will inquire my manager’s opinion about personal matter that is not related to work”. Originally the IPI items were rated with 7 alternatives ranging from 1 (definitely not a reason) to 7 (definitely a reason) from complying. In this study, the IPI Items were rated on a 6-point likert scale: (6) strongly agree, (5) agree, (4) moderately agree, (3) disagree, (2) moderately disagree, and (1) strongly disagree. IPI was translated using backto-back translation procedure. Decentralization and back-to-back translation of IPI items The In the decentralization process, the original measurement was changed before it was adapted and back-to-back translated. [20] and [21] assert that two different bilingual language experts has to be used in the back-to-back translation process. Thus, in this process, an expert Malay translator translated the 44 items to the Malay language and another English expert retranslated the translated items into the English language without having seen the original items. After that, the quality of the language translation was observed in terms of how accurately the back translated measurement agrees with the original version. Then, the back translated items were discussed and verified with Human Resources Managers and clerical staff from the GLCs (to ensure the suitability of all items), 2 academicians who are experts in organizational communication field to get feedbacks on the appropriateness of items adapted and translated in measuring social B.
power. This stage is crucial to guarantee content and face validity of all items used in the study. Several amendments were made to the items based on the feedbacks given. C. Analytic Procedures Data were analyzed using Statistical package for Social Science (SPSS). The reliability and initial evidence of validity were reported based on results from Cronbach’s alpha reliability and EFA. The EFA on the latent construct was carried out to determine if the responses gathered can be grouped according to items in each of the hypothesized dimensions. Following [22] and [19], [23], [24], EFA using principal axis factoring with varimax rotation and a priori data of 11 factors was conducted to analyze factor structure of the construct. The cutoff point of 0.5 was used as the threshold to ensure practical significance for further analysis [23]. To analyze and generate sufficient variance among respondents for subsequent statistical analysis as well as identifying and confirming the underlying structure of the 44 items of IPI, a sample size of 101 executives working directly under Human Resource Managers (HR Managers) was used as a pilot study. The EFA purpose of this study was conducted to examine the stability of the factor structure and provide information that would facilitate the refinement of a new IPI measure.
IV. RESULTS AND DISCUSSIONS Reliability The EFA results are shown in Table 1. The 44 items of the IPI were subjected to Principal Components Analsyis (PCA) using SPSS. Prior to performing PCA, the suitability of data for factor analysis was assessed. Inspection of the correlations matrix revealed the present of many coefficients of 0.1 and above. The Kaiser-Mayer Olkin of Sampling Adequacy (KMO) value was 0.451, below the recommended value of 0.6 (Kaiser, 1974). The Bartletts’s Test of Sphericity for the 44 items correlation matrix was not significant (p>0.000). The overall internal consistency reliability for the IPI scale was 0.551. A.
International Conference on ArtS and Sciences – 2014 (IJAC’14)
The analysis indicated that some of the item did not group together. Looking at the communalities table, items RP5, RP6, RP8, CP13, CP14, CP15, CP16, LP21 – LP27, LP29 – LP32, IP40 – IP44, illustrated loadings below the threshold of 0.50, therefore, researchers decided to delete the abovementioned items. Based on reliability results, it were indicated that all items (4 items) under dimension of informational power, 4 items under legitimacy-reciprocity, 3 items under legitimacydependence, 4 items under legitimacy-equity, 3 items under personal reward, and 4 items under personal coercion had poor factor loadings and cronbach’s alpha (below the threshold of 0.50). Also the scree plot suggested that five factors should be retained. Again, looking at the commonalities table, items with communalities less than .40 were flagged for further investigation. The factor analysis was rerun until a satisfactory result was achieved. After the purification process, a total number of 21 items were deleted from the original IPI scale, leaving only 23 items, and 2 newly items (EPN) (RfPN) were reained. The new IPI scale shows (a) 4 items for reward impersonal, and 1 item for personal reward overlapped with reward impersonal and \ was labeled as Reward Power; (b) 4 items for coercive impersonal and 1 item under impersonal coercion was also combined and labeled as Coercive Power; (c) 4 items under the factor of legitimate position were maintained and combined with 1 item left under legitimacydependence was labeled as Legitimate Power; (d) all 4 items for expert power were maintained, and 1 new item added (based on previous scale in the literature) (EPN) and labeled as Expert Power; and finally 4 items in referent power constructs were maintained, adding 1 item (RfPN) based on the previous scale in the literature). This brings altogether 25 items after the deletion of 21 items in IPI and 2 new items being added under expert and referent powers respectively. The process of scale purification in this initial stage reduced the number of items from 44 to 23 items and addition of 2 new items. Therefore, a total of 25 items remained for the new IPI scale. It was
TABLE I: THE PSYCHOMETRIC PROPERTIES OF THE ELEVEN: DIMENSIONS OF THE IPI, 2013 Factor Reward Impersonal
Personal Reward
Coercive Impersonal
Personal Coercive
Legitimate/ Position
Legitimate/ Reciprocity
Legitimate/ Dependence
Legitimacy/ Equity
Expert Power
Newly added Referent Power
Newly added Information Power
Item
Loadings
RP1 RP2 RP3 RP4 RP5 RP6 RP7 RP8 CP9 CP10 CP11 CP12 CP13 CP14 CP15 CP16 LP17 LP18 LP19 LP20 LP21 LP22 LP23 LP24 LP25 LP26 LP27 LP28 LP29 LP30 LP31 LP32 EP33 EP34 EP35 EP36 *EPN RfP37 RfP38 RfP39 RfP40 *RfP IP41 IP42 IP43 IP44
0.750 0.725 0.721 0.810 0.443 0.431 0.710 0.353 0.743 0.821 0.859 0.893 0.421 0.399 0.441 0.801 0.623 0.621 0.637 0.648 0.337 0.452 0.412 0.399 0.382 0.297 0.288 0.711 0.448 0.387 0.422 0.367 0.719 0.732 0.694 0.801 0.620 0.721 0.799 0.754 0.717 0.821 0.344 0.444 0.234 0.239
Common alities 0.726 0.795 0.778 0.722 0.561 0.467 0.757 0.601 0.730 0.798 0.857 0.848 0.491 0.512 0.444 0.799 0.566 0.611 0.621 0.639 0.554 0.546 0.435 0.673 0.434 0.561 0.402 0.663 0.435 0.455 0.568 0.459 0.712 0/785 0.710 0.822 0.769 0.655 0.732 0.753 0.774 0.893 0.482 0.421 0.429 0.495
MSA
0.701
0.716
0.513
0.555
0.739
0.752
0.432
0.498
0.759
0.820
0.412
0.419
0.399
0.398
0.499
0.532
0.790
0.792
0.782
0.799
0.431
0.511
*EPN *RfP new item added
concluded that the first round of factor analysis (as shown in Table 1) and internal consistency
International Conference on ArtS and Sciences – 2014 (IJAC’14)
reliability, mean and standard deviation for the total score and each subscale of 44 items showed a very poor Cronbach’s alpha values (ranges within 0.511 and 0.820) for all the 11 subscales. After the deletion of 21 items and addition of 2 items, the same procedure of EFA was carried out again to confirm the reliability and validity of the new constructs of IPI. This procedure was carried out with the new 25 items of IPI(M) to determine if the responses gathered can be grouped according to items in each of the new dimension being developed. Researches decided to add (M) after IPI to differentiate IPI from [3] that probably can be used to measure social power in the Malaysian organizational context. B. Consruct Validity
Construct validity refers to whether a scale measures or correlates with a theorized psychological construct. To examine IPI competencies scale the EFA with varimax rotation was performed to identify and confirm the underlying structure of the items. Initially, the suitability of the data for factor analysis was explored. A second round of EFA analysis was conducted to confirm the underlying structure of the new 25 items IPI(M) scale. Table 2 shows the final round of EFA analysis which confirmed 5 dimensions of IPI(M) namely Reward, Coercive, Legitimate, Expert, and Referent Power. The new IPI(M) scale that consist of 25 items shows that the KMO measure of sampling adequacy has increased to 0.899, exceeding the recommended value of 0.6 [25], indicating that the sample size was adequate to factor analyze the 25 items. The Chi-Square value of Bartlett’s Test of Sphericity, for the 25 items correlations matrix was highly significant (p