García-Arreola Oscar, Unda-Rojas Sara, Hernández-Toledano Rosa Alejandra, Tovalin-Ahumada Horacio - PDF

Description
Psychology Research, April 2016, Vol. 6, No. 4, doi: / / D DAVID PUBLISHING Psychometric Behavior of the Short Version of Effort-Reward Imbalance Questionnaire in a Sample

Please download to get full document.

View again

of 11
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information
Category:

Art

Publish on:

Views: 10 | Pages: 11

Extension: PDF | Download: 0

Share
Transcript
Psychology Research, April 2016, Vol. 6, No. 4, doi: / / D DAVID PUBLISHING Psychometric Behavior of the Short Version of Effort-Reward Imbalance Questionnaire in a Sample of 395 Mexican Teachers * García-Arreola Oscar, Unda-Rojas Sara, Hernández-Toledano Rosa Alejandra, Tovalin-Ahumada Horacio UNAM (Universidad Nacional Autónoma de México), Ciudad, México This study examined the psychometric properties of the short version of the ERI (Effort-Reward Imbalance) questionnaire with a total of 395 high school teachers (50.1% male, 49.9% female). The reliability of the questionnaire was obtained by an analysis of internal consistency, the validity of the instrument by Exploratory Factor Analysis and construct by CFA (Confirmatory Factor Analysis). Finding Alfa s values greater than 0.60; The CFA present good indices of fit ( = 173,451(72), GFI = 0.94, RMSEA = 60, and λ weight of 46-99). In conclusion the shorter version of ERI questionnaire has acceptable psychometric properties in this teacher s population. Keywords: ERI (Effort-Reward Imbalance), teachers, CFA (Confirmatory Factor Analysis), work, stress Introduction At this point in the Americas, the occupational health field faces the challenge to counteract the actual labor trends based in the intensification of the job demands, flexibility of the employment conditions and an alarming increase in the number of workers that develop serious physical and mental diseases. To improve the occupational health conditions, the identification and surveillance of psychosocial risks factors at workplaces is necessary to protect to the employees. Therefore, it s crucial the locally validate instruments to assess the current status of the psychosocial risk factor at work, in this way, the capacities of the occupational health in this area are strengthen (Muñoz et al., 2000). This reflects the need for practical and sensitive tools that focuses on occupational psychosocial risk factors prevention and control (Charria, Sarsosa, & Arenas, 2011). In regard to the detection and control of psychosocial risks, there are two models that have great relevance the Demand-Control model and ERI (the Effort-Reward Imbalance) model (Aguado, Bátiz, & Quintana, 2013; Charria et al., 2011; Feldman & Blanco, 2012). From these models were developed the most used questionnaires to assess work stressors: the Job Content Questionnaire (JCQ) (Karasek & Theorell, 1990) and the Effort Reward Imbalance Questionnaire (ERI) (Siegrist, 1996). * Acknowledgements: This research was conducted within the project UNAM-DGAPA-PAPIIT IN306514, the authors are grateful for the support. García-Arreola Oscar, B.S., Carrera de Psicología, Facultad de Estudios Superiores Zaragoza, UNAM (Universidad Nacional Autónoma de México). Unda-Rojas Sara, Ph.D., Carrera de Psicología, Facultad de Estudios Superiores Zaragoza, UNAM (Universidad Nacional Autónoma de México). Hernández-Toledano Rosa Alejandra, B.S., Carrera de Psicología, Facultad de Estudios Superiores Zaragoza, UNAM (Universidad Nacional Autónoma de México). Tovalin-Ahumada Horacio, Ph.D., Especialidad de Salud en el Trabajo, Facultad de Estudios Superiores Zaragoza, UNAM (Universidad Nacional Autónoma de México). 210 PSYCHOMETRIC BEHAVIOR OF THE SHORT VERSION OF EFFORT-REWARD The ERI model sees the stress as the product of a socially organized process, related to the imbalance between effort (extrinsic and intrinsic) and the reward at work. Stress exists in conditions of imbalance between the effort and reward, which can produce a state of anxiety and stress, with an adverse health effect (Luceño, García, Rubio, & Díaz, 2004; Siegrist, 1996) (see Figure 1). Years later, Siegrist (1999) established the intrinsic component as an independent variable, thus combine the information on the demands and rewards at work (extrinsic component) with information on the personal characteristics, determined by a strong commitment and a high need for approval (intrinsic component) that influences the perception of the extrinsic component. In this way, the model contemplates the complexity of individual stressful experiences and those related to employment situation (see Figure 2) (Macias et al., 2003; Siegrist, 2008). Figure 1. The initial ERI model (Siegrist 1996, p.30). Figure 2. Actual ERI model (Siegrist, 1999, p. 40, as mentioned in Vagchel, de Jonge, Bosma, & Schaufeli, 2005). The operationalization of the ERI model was carried out through the creation of a self-report questionnaire, the original version contains 22 items that has been translated into different languages. The ERI Spanish version was validated in different studies, Díaz and Feldman (2010), Gómez (2010), Macias et al. (2003), and in Mexico by Arias (2015) and Camacho, García, Noriega, Escobedo, and Juárez (2015), and the version supports the validity and reliability of the questionnaire and model. PSYCHOMETRIC BEHAVIOR OF THE SHORT VERSION OF EFFORT-REWARD 211 Lately, Siegrist developed a 16 items version, this version facilitates the application, and have acceptable alpha scores in all the dimensions, Effort, Reward and Over commitment (α = 0.74, α = 0.79, and α = 0.79, as listed) (Siegrist, Wage, Pühlhofer, & Wahrendorf, 2009; Siegrist, Li, & Montano, 2014). This shorter version of the ERI questionnaire is considered a useful tool for epidemiological studies regarding the effects of work in health in globalized context. A validation conducted by Leineweber et al. (2010) with a sample of 4,771 workers, obtained acceptable Cronbach s alpha coefficients for the Effort (α = 0.80), Reward (α = 0.84), and Overcommitment (α = 0.85), and its confirmatory analysis showed good fit indices ( = , df = 98, = , GFI = 0.94, AGFI = 0.91, RMSEA = 0.06, CFI = 0.93 and CAIC = ). A Chinese validation (Li et al., 2012) with a sample of 1,916 workers obtained a Cronbach s alpha coefficients of Effort, Reward and Overcommitment of 0.86, 0.72, and 0.73 respectively, and the confirmatory analysis showed good fit indices ( = , GFI = 0.95, AGFI = 0.93, RMSEA = 0.06, CFI = 0.90 and CAIC = ). In Japan, Kurioka, Inoue, and Tsutsumi (2013) contrasted the two versions of the questionnaire, with a sample of 1,489 Japanese workers, the authors concluded that the short version of the ERI is concise and psychometrically valid. However, there was a disagreement in the classification as high-risk individual using the original cut-off point in the original and shorter versions. For this reason, the author suggested for the shorter version a different cut-off point value of ER 1.4. The instrument has been tested in several studies and it serves as a powerful predictor of psychosomatic symptoms (de Jonge, Bosma, Peter, & Siegrist, 2000; Tsutsumi & Kawakami, 2004) and shows a significant relationship between Effort-Reward imbalance and the occurrence of cardiovascular diseases (increased heart rate and blood pressure), and the disrupted secretion of stress hormones (cortisol, epinephrine, and norepinephrine) (Fernández, Siegrist, Rödel, & Hernández, 2003; Siegrist, 1996, 2010). However, to guarantee the quality of measurements, this instrument must be subject to a validation process (Carvajal, Centeno, Watson, Martínez, & Sanz, 2011). This study aim was to analyze the psychometric properties of the shorter version of the ERI in a group of Mexican teachers. Methods Sampling and Data Collection The questionnaire was distributed by to 4,000 teachers, obtaining a response rate close to 10%. Participants were informed about the research project, their participation was voluntary, they signed and informed consent letter, and the confidentiality of their data was guaranteed. The participants in this study was a non-random sample of 395 teachers, 198 men (50.1%) and 197 women (49.9%) from ten different schools from a Mexican university. The teachers had a mean age of years old, (SD = 10.57, Range years old), a mean seniority of years (SD = 11.63, range 1-50 years), and 63.5% have a couple. Most of them had a bachelor degree (58.2%), followed by a master degree (36.5%), and a Ph.D. (5.3%). Regarding their type of contract, 76.5% were part-time and 23.5% of full time teachers. A 58.8% of them were in a productivity program. Instruments and Procedures The Spanish version of the ERI short version used in this study was from the official ERI webpage (Siegrist, 2012): (1) To get the information the following questionnaires were applied; 212 PSYCHOMETRIC BEHAVIOR OF THE SHORT VERSION OF EFFORT-REWARD (2) A questionnaire to get demographic and work environment information. The ERI short version questionnaire: (1) To evaluate the extrinsic component of the model, there are 10 items, to get information from the worker s perspective on Effort (3 items, 1-3; range, 3-12 points) and professional Reward (7 items, 4-10; range 7-28 points); (2) To explore the intrinsic component, there are 6 items, to know the degree of involvement experienced by the worker in his job, represented by a unique factor called Overcommitment (6 items, range 6-24 points). The imbalance between Effort and Reward (ER) can be calculated with the logarithm of, where: (1) E is the score obtained in Effort; (2) R is the score in Reward; and (3) k is a correction factor to adjust an unequal number of elements between Effort and Reward. If an Effort Item is equivalent to one from Reward, then k is: (. The result can be interpreted as follows: if ER = 1 is assumed that the person reports the same level of Effort and Reward; if ER 1 there is less Efforts than Rewards, and if ER 1 means that the person reports more Efforts than Rewards (Siegrist, Li, & Montano, 2014). Data Analysis The SPSS and AMOS software was used for the statistical analysis. Internal consistency analysis was performed for each item with Cronbach s alpha criteria scores ranging between as suggested by Oviedo and Camos-Arias (2005). For the validity of the instrument, a EFA was performed by the method of Main Components using a rotating VARIMAX, requesting three factors as theory commands and three without a minimum number of factors, one general and two other for sex. A Bartlett sphericity test and the index of sampling adequacy Kaiser-Mayer-Olkin (KMO) were carried out, expecting acceptable values p 0.05 and KMO 0.70 respectively (Pérez & Medrano, 2010). Subsequently to evaluate the validity of the model a CFA with an assembly of structural equations with a maximum likelihood method, measures of absolute, parsimonious and incremental adjustment were used as suggested by Lévy, Martín, and Román (2006). Results Internal Consistency In the analysis of internal consistency by item, the highest measure appeared in item seven My job security is poor with 2.84, while item three showed the lowest value with 1.76 Over the past few years, my job has become more and more demanding (see Table 1). All CHI (corrected homogeneity indexes) were greater than 0.40, with the exception of item two within the Effort dimension I get many interruptions and disturbances while performing my job (0.35); the highest homogeneity index belonged to Work rarely lets me go, it is still on my mind when I go to bed (0.72) and As soon as I get up in the morning I start thinking about work problems (0.68). PSYCHOMETRIC BEHAVIOR OF THE SHORT VERSION OF EFFORT-REWARD 213 The results of Cronbach s alpha showed that there is an acceptable reliability in Reward (0.82) and Overcommitment (0.80), but Effort (0.67) scored below the recommended criteria of 0.70 (Oviedo & Campo-Arias, 2005). For the Effort dimension, almost all the items had reliability indexes above 0.3, and most of them contribute to the internal consistency factor, but the item two I have many interruptions and disturbances while performing my job has a low reliability. If item two is removed, the dimension s Cronbach s alpha score increases to an acceptable value of Table 1 Analysis of Discrimination ERI Reagents Average SD Homogeneity corrected Index from reliability Effort Alpha = ERI1 I have constant time pressure due to a heavy work load ERI2 I have many interruptions and disturbances while performing my job ERI3 Over the past few years, my job has become more and more demanding Reward Alpha = Esteem Alfa = ERI4 I receive the respect I deserve from my superior or a 2.46 respective relevant person ERI8 Considering all my efforts and achievements, I receive the respect and prestige I deserve at work Job security Alpha = ERI6 I have experienced or I expect to experience an undesirable change in my work situation ERI7 My job security is poor Job promotion Alpha = ERI5 My job promotion prospects are poor ERI9 Considering all my efforts and achievements, my job promotion prospects are adequate ERI10 Considering all my efforts and achievements, my salary/income is adequate Overcommitment Alfa = 0.80 OC1 I get easily overwhelmed by time pressures at work OC2 As soon as I get up in the morning I start thinking about work problems. OC3 When I get home, I can easily relax and switch off work. OC4 People close to me say I sacrifice too much for my job. OC5 Work rarely lets me go, it is still on my mind when I go to bed. OC6 If I postpone something that I was supposed to do today I ll have trouble sleeping at night. Alpha without reagent 214 PSYCHOMETRIC BEHAVIOR OF THE SHORT VERSION OF EFFORT-REWARD Table 2 presents the correlation among the ERI dimensions, the stronger association was observed between Effort and Overcommitment (0.566). There was a significant correlation between all of dimensions. There was a negative significant association among Reward and Effort and Overcommitment. Table 2 Pearson Correlations: ERI Model Dimensions Variables Effort 2. Reward ** 3. Overcommitment ** ** Note. ** Correlation is significant at the 0.01 level. Table 3 shows the results of the factorial analysis with the requirement of three factors, according to the theory, as well as three analysis without a fix number of factors, one global and two per gender suggested in Pérez and Medrano (2010), for a sample bigger than 300. Table 3 Factor Analysis of Overall and Sex Dimension Items Effort Esteem Job security Job promotion Overcom mitment Fixed number of factors No fixed number of factors Overall Women Men ERI ERI ERI ERI ERI ERI ERI ERI ERI ERI OC OC OC OC OC OC The factorial analysis with the requirement of three factors behaves as expected by the model with the exception of the first item of the Overcommitment dimension in the Effort scale. The analysis without a fix number of factors shows four factors and the Job security scale does not group with the Reward dimension. The first component includes all items of Effort and one from Overcommitment (α = 0.75), the second component includes the Esteem and Job Promotion dimensions (α = 0.80), the third component Job security (α = 0.75) and the last component was integrated by five items of Overcommitment (α = 0.78). PSYCHOMETRIC BEHAVIOR OF THE SHORT VERSION OF EFFORT-REWARD 215 In the case of woman, the factors had a similar pattern than the global analysis. For men, the organization and composition and weight of the components changes, the first component includes all Effort and five items of Overcommitment, the second component includes Esteem, Job promotion and one item from Job security and the third component includes items from Overcommitment and Job security. Most of the items in the factorial analysis results were properly organized with acceptable scores. KMO = and = ; p 0.000, that said the factorization of the correlation matrix is feasible. When performing the CFA with the method of maximum likelihood, the results of the a priori model proposed by Siegrist (2009) presents in the dimension of Effort an inadmissible commonality (case Heywood ) (McDonald, 1985). The same analysis was performed by the generalized least squares method and the weighted least squares method that according to Lloret-Segura, Ferreres-Traver, Hernández-Baeza, and Tomás-Marco (2014) reduces the probability to find Heywood cases. At present, the same problem decided accommodate the dimensions according to EFA, for this analysis the questionnaire was purged removing those items with the highest rates of change in the regressions, located in the OC2 and OC4 reagents. When suppressing the analysis reagents by the method of maximum likelihood, the results were favorable (see Table 4). The adjustment index chi-square ( ) it was significant, although this result was expected given the size of the sample. However the GFI and RMSEA were acceptable values, indicating a good degree of accuracy in the model. The incremental adjustment NFI and CFI gain acceptable value above 0.90 in both cases. Finally, the fit indices PGFI and normalized Chi-square X 2 /gl, shows a low parsimony in the model. Finally, in the second order model (see Figure 4), the lowest weight standardized regression started off item 2 I have many interruptions and disturbances in the performance of my job, λ = 0.46 and the highest point is found between construct ERI and Effort λ = Table 4 Adjustment Indices Absolute fit indices Incremental fit indices Parsimony fit indices Weight from regressions Standardized GFI RMSEA NFI CFI PGFI X 2 /gl Weight (λ) (72) p 216 PSYCHOMETRIC BEHAVIOR OF THE SHORT VERSION OF EFFORT-REWARD Figure 4. Second order structural model. Conclusions and Discussion According to the analysis, the reliability of the instrument was found to have a good internal consistency similar to that reported by other countries (Díaz & Feldman, 2010; Gómez, 2010; Macías et al., 2003; Siegrist, Li, & Montano, 2014). In contrast to other validations of the short version, the values reported in Effort do not fit within the acceptable limits, just like those reported in the Chinese validation (0.67), contrasting with the Swiss and German validations, where alpha values stood 0.80 and 0.74 respectively (Li et al., 2012; Leineweber et al., 2010; Siegrist, Wage, Pühlhofer, & Wahrendorf, 2009, respectively). It is important to point out that item 2 I have many interruptions and disturbances while performing my job contributed to inconsistencies by not backing the Alfa in the Effort dimension, its elimination resulted in an increase alpha of 0.73, and this would result in an acceptable Alfa, similar to the German validation. PSYCHOMETRIC BEHAVIOR OF THE SHORT VERSION OF EFFORT-REWARD 217 On the other hand, the EFA values of interrelation are greater than 0.40 which shows good consistency in size, although the item I get easily overwhelmed by time pressures at work loaded of Effort rather than the Overcommitment factor in all EFA. Similar to that found on petrochemical industry workers in Korea (Eum et al., 2006) and in Chinese health care workers (Li, Yang, & Cheng, 2005). The results confirmed the hypothesis of Li et al. (2005) who point out that there are cultural variations in the perception of stressful experience in the work. Yet, it may also be due to specific conditions of the teaching profession. The gender differences found, shows the necessity of analysis, considering the work culture and the differences between women and men. In this study the men have a configuration of the first ERI model. Reward dimension in its original version is made up of three components (Esteem, Job security and Job promotion) (Li et al., 2012; Leineweber et al., 2010; Siegrist et al., 2009; Siegrist et al., 2014). That contrasts with the two dimensions found, one composed of Esteem and Job promotion and the other formed by the dimension of Job Security. This distinction confirms the statement by Zurlo, Pes and Siegrist (2010), who in a sample of 673 Italian teachers found that the ERI model has a different configuration of the Reward dimension, in one hand, the immaterial component Esteem reward and in the other hand, a material component, where salary and career reward, originating
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks