EN ES
Vol. 31. Num. 3. 2015. Pages 187-200

The Work Design Questionnaire: Spanish version and validation

[Validación española del Work Design Questionnaire]

Bayona, Jaime Andrés Caballer, Amparo Peiró, José-María Pontificia Universidad Javeriana, Bogotá, Colombia

Abstract

The purpose of this study is to validate the Spanish version of the Work Design Questionnaire (WDQ; Morgeson & Humphrey, 2006). Employees from three Colombian samples completed the questionnaire (N = 831). Confirmatory factor analyses revealed a 21-factor structure ( 2 /df ratio = 2.40, SRMR = .06, RMSEA = .04, CFI = .90) with adequate levels of convergent and discriminant validity. Additional support for construct validity was found from significant differences among different occupational groups (professional and nonprofessional, health-focused, commercial, and manufacturing workers). Furthermore, knowledge, social, and work context characteristics showed incremental validity over task characteristics on job satisfaction and perceived performance. Possible interpretations of these relationships are offered. It is concluded that the study provides evidence for the validity of a Spanish version of the scale, and presents further support for the generalization of the 21-factor structure of work design characteristics in different cultural settings. 

Resumen

El propósito de este estudio es validar la versión española del Work Design Questionnaire (WDQ; Morgeson y Humphrey, 2006). Tres muestras de empleados colombianos completaron el cuestionario (N = 831). El análisis factorial confirmatorio reveló una estructura de 21 factores (razón 2 /gl = 2.40, SRMR = .06, RMSEA = .04, CFI = .90) con adecuados niveles de validez convergente y discriminante. Se encontraron diferencias significativas entre diferentes grupos ocupacionales (profesionales, no profesionales, trabajadores de la salud, comerciales y de producción). También se encontró que las características del conocimiento, sociales y contextuales aportaron validez incremental sobre la satisfacción laboral y el desempeño percibido. Se ofrecen posibles interpretaciones de estas relaciones. Se concluye que el estudio proporciona evidencia suficiente sobre la validez de la versión española de la escala, lo que presenta más apoyo para la generalización de la estructura del modelo de características del trabajo de 21 factores en diferentes contextos culturales. 

Work Design “describes how jobs, tasks, and roles are structured, enacted, and modified, as well as the impact of these structures, enactments, and modifications on individual, group, and organizational outcomes” ( Grant & Parker, 2009 , p. 319). From the early studies on task attributes ( Turner & Lawrence, 1965 ) to the interdisciplinary approach to work design ( Campion, 1988 ), it has been a demand, from both scientists and practitioners, to have a valid and reliable instrument assessing work characteristics in organizational settings. During the last 30 years questionnaires such as the Job Diagnostic Survey (JDS; Hackman & Oldham, 1975) and the Multimethod Job Design Questionnaire (MJDQ; Campion, 1985 ) have been developed to assess work design characteristics; however, these instruments generally have suffered from two drawbacks: (a) questionable psychometric properties related with the low internal consistency of the JDS ( Kulik, Oldham, & Langner, 1988; Taber & Taylor, 1990 ) and problems with the factor structure of the MJDQ ( Edwards, Scully, & Brtek, 1999 , 2000 ); and (b) a mismatch between the work characteristics assessed by the instruments and the real characteristics presented in nowadays organizational settings, that is represented in a shift from manufacturing economies to service and knowledge economies that had altered the nature of work in organizations ( Grant & Parker, 2009 ). Due to these limitations, Morgeson and Humphrey (2006) developed the Work Design Questionnaire (WDQ), that presents both high reliable psychometrics and takes into account current models of work design ( Grant, Fried, & Juillerat, 2010; Humphrey, Nahrgang, & Morgeson, 2007 ).

This need for a valid and reliable instrument is especially relevant in non-English speaking countries, where work dynamics have changed during the last 20 years and old work design instruments are no longer appropriate for these new organizational settings. Thus, the purpose of the present study is to validate a Spanish version of the WDQ with a sample of Colombian workers.

Work Characteristics Assessment

From the early work of Turner and Lawrence (1965) , work characteristics have been assessed mainly through self-report questionnaires that ask workers to rate their personal evaluation of the presence of certain work attributes. Using this approach, two major work design questionnaires have been developed: the Job Design Survey (JDS) and the Multimethod Job Design Questionnaire (MJDQ).

Richard Hackman and his colleagues developed the JDS as an instrument to assess the job characteristics model (JCM) ( Hackman & Lawler, 1971; Hackman & Oldham, 1975, 1976 ), which has been the standard model in work design for both academics and practitioners during the last 40 years. The JDS is a self-reporting instrument meant to diagnose the motivational properties of a job prior to a redesign procedure. The major contribution of the JCM and JDS was that it established that core job characteristics are associated with favorable attitudinal and behavioral reactions ( Grant et al., 2010 ). However, the main criticisms to JCM were: (a) the treatment of within-person relations as person-situation relations, (b) the model structure, due to some inconsistences in the role of the moderator and mediators, (c) the small subset of characteristics d in the model, (d) concerns about the convergent and divergent validity of the JDS, and (e) the theoretical and mathematical justification of the composite job characteristics index ( Fried & Ferris, 1987; Johns, Xie, & Fang, 1992; Roberts & Glick, 1981 ).

Taking into account some of these criticisms, a new model of work design emerged: the interdisciplinary model of job design ( Campion, 1988; Campion & Thayer, 1985 ) which aimed to develop a new taxonomy of work design that d 48 different job characteristics with a 48-item questionnaire. The major strengths of this approach were: (a) the inclusion of new work characteristics that were relevant to the work context and (b) the discovery that different job design approaches influence different outcomes. On the other side, the major weakness of the interdisciplinary model lay in the psychometric proprieties of the MJDQ, especially the construct validity, since every dimension was assessed by only one item ( Edwards et al., 1999).

From these previous models, Frederick Morgeson and Stephen Humphrey developed an inductively generated collection of work design characteristics that integrated the work design literature into four major work characteristics: (a) Task Characteristics, which work scheduling autonomy, decision-making autonomy, work methods autonomy, task variety, task significance, task identity, and feedback from job; (b) Knowledge Characteristics, which job complexity, information processing, problem solving, skill variety, and specialization; (c) Social Characteristics, which social support, initiated interdependence, received interdependence, interaction outside the organization and feedback from others; and (d) Work Context Characteristics, which ergonomics, physical demands, work conditions, and equipment use ( Morgeson & Humphrey, 2006 ). This taxonomy integrated some elements of previous models but d new characteristics that are present in today's organizations (i.e., knowledge characteristics that reflect the current knowledge work and social characteristics that reflect the emphasis on service organizations that rely more deeply on social interactions).

The construction of the WDQ was developed through five stages: (a) review of job characteristics in the literature and grouping of the resulting characteristics into a 21 characteristics proposal, (b) literature review to search items that evaluate each job characteristic, (c) adapting items and creating new items for the 21 characteristics proposal, (d) statistical analyses of the 21 job characteristics proposal using confirmatory factor analysis (CFA), and (e) construct validity analyses using O*NET and checking relationship between occupations and various outcome measures ( Morgeson & Humphrey, 2006 ). The results of this procedure gave support to a 21-factor structure with a high reliability and convergent and discriminant validity, which in turn resolved two of the major criticisms of previous work design instruments: the limited number of job characteristics considered (JDS) and the weak psychometrics (MJDQ).

Work Design in Spanish Speaking Countries

All preceding models were developed within the North American context, with research on work design in Spanish speaking countries dealing mainly with: (a) the validation of work design instruments in their cultural settings or (b) the use of a work design instrument as a measure within a broader research.

The research on work design in Spanish speaking countries from a validation perspective s a couple of JDS validations ( Dávila & Chacón, 2003; Fuertes, Munduate, & Fortea, 1996; Martínez-Gómez & Marín-García, 2009 ) which confirmed the 5-factor dimension structure but with some reliability problems, especially in the skill variety, autonomy, and identity dimensions. From the second perspective, the work design research on those countries was particularly associated with the use of JCM. In Spain there had been a number of studies using the JDS, as in a study of burnout, organizational climate, and work motivation ( Boada, Vallejo, & Agulló, 2004 ), in which three out of five JCM dimensions were associated with different burnout outcomes (autonomy, skill variety, task significance). Other research studied the influence of communication skills on work teams management ( Ramis, Manassero, Ferrer, & García-Buades, 2007 ), in which no direct effect of job characteristics was associated with leader communication skills. Finally, a study on the redesign of tasks in the Spanish automotive industry concluded that all JCM dimensions were related with attitudinal outcomes (especially autonomy and feedback), but not with any performance outcomes ( Osca & Urien, 2001 ). In Latin America, research on work design was more limited: two studies, including the JDS, were conducted in Perú, one that sought to explore the utility of the socio-technical systems theory in that country, which reported a significant influence of feedback on the degree of technology implementation ( Salas & Glickman, 1990); the other study, by Solf (2006) , used a section of JDS (employee growth need strength) to investigate labor intrinsic motivation and personality in a sample of Peruvian workers.

These studies in Spanish speaking countries indicate that there is an interest in the design area, even though the tools available offered a limited range of job characteristics and the psychometric specifications were not the most appropriate nowadays. Taking these problems into account, the purpose of the present research is to bridge this gap by adapting the WDQ to the Spanish language, which will offer researchers and practitioners a valid and reliable instrument to work with Spanish-speaking workers in the area of work design.

Hypotheses Development

In this paper we describe the adaptation process of the WDQ into Spanish. We tested the psychometric properties of the adapted version through a variety of means. First, we conducted a series of CFAs to confirm the factor structure of the Spanish adaptation; second, we further examined the psychometric properties of the scale by testing its capacity to differentiate across occupations; third, we explored the relations of major work characteristics with job satisfaction and perceived performance; finally, we examined the incremental validity of work characteristics for job satisfaction and perceived performance.

In the original article about WDQ, five different factor structure models were tested: (a) a 4-factor model examines the four broad categories of work characteristics (task, knowledge, social, and context); (b) an 18-factor model examines each work characteristic; (c) a 19-factor model separates interdependence into its received and initiated components; (d) a 20-factor model separates autonomy into its three components, which s autonomy in work scheduling, decision making, and work methods; (e) finally, a 21-factor model separates both interdependence and autonomy into the identified components. Following the results from the original English version ( Morgeson & Humphrey, 2006) and the German (Stegmann et al., 2010) and Italian ( Zaniboni, Truxillo, & Fraccarolli, 2013 ) validations of the WDQ, it is expected that the Spanish version of the WDQ will fit into a 21-factor model. Hypothesis 1

The Spanish WDQ version represents a 21-factor structure.

In order to validate the WDQ, it is important that the Spanish version could detect differences across occupations according to their job and role contents, because the original WDQ is aimed at differentiating between jobs, and thus, the Spanish version should be able to differentiate between different classes of jobs. In the original WDQ validation article, four different occupational groups were compared in different work characteristics (professional, non-professionals, human-life occupations, and sales occupations). In line with this procedure, we examined the differences in work characteristics in four groups. First, we expected that “jobs in professional occupations would be higher on both the broad set of knowledge characteristics and the three components of autonomy than jobs in nonprofessional occupations, because professional occupations generally involve complex, non-routine work that s flexible and adaptive behavior where higher levels of autonomy are present” ( Morgeson & Humphrey, 2006 , p. 1328). Second, we expected “jobs in nonprofessional occupations, compared with those in professional occupations, to be higher on physical demands and lower in the quality of work conditions because these jobs generally involve more physical exertion in less than optimal work environments” ( Morgeson & Humphrey, 2006 , p. 1328). Third, we expected jobs in health related occupations to be higher on task significance because behavior in these occupations directly affect human lives ( Morgeson & Humphrey, 2006 ). Finally, we expected jobs in commercial occupations to be higher on interaction outside the organization because sales occupations are specifically focused on providing products and services to others ( Morgeson & Humphrey, 2006 ). According to this rationale we formulate the following hypotheses. Hypothesis 2a

Professional occupations will have higher levels of knowledge and autonomy characteristics than nonprofessional occupations.

Hypothesis 2b

Nonprofessional occupations will have higher levels of physical demands and less positive work conditions than jobs in professional occupations.

Hypothesis 2c

Jobs in health occupations will have higher levels of task significance than manufacturing occupations.

Hypothesis 2d

Jobs in commercial occupations will have higher levels of interaction outside organization than manufacturing occupations.

Research on work design has found a positive relationship between task characteristics and attitudinal (i.e., job satisfaction) and behavioral (i.e., performance) outcomes ( Fried & Ferris, 1987; Hackman & Oldham, 1980 ). In the original WDQ validation article, task and knowledge characteristics were compared to job satisfaction. In addition, some research has stated that the new work design characteristics (e.g., knowledge and social ones) will a similar relation to job satisfaction and perceived performance ( Grant et al., 2010; Grant & Parker, 2009 ). Following this rationale we expected that task, knowledge, and social characteristics would be related to both job satisfaction and perceived performance. Hypothesis 3

Task, knowledge and social characteristics will be positively related to job satisfaction ( Hypothesis 3a) and perceived performance (Hypothesis 3b).

In the original WDQ validation article, social support was expected to incrementally predict satisfaction beyond task characteristics; however, due to changes in the nature of work that emphasizes more knowledge and service jobs than industrial ones ( Grant & Parker, 2009 ), and due to the characteristics of Latin-American countries in which social relations are highly valued ( Hofstede, 2001; House, Hanges, Javidan, Dorfman, & Gupta, 2004 ), it is expected that both knowledge and social characteristics will influence job satisfaction and perceived performance in an incremental rate. In addition, although some authors considered that context characteristics explained little variance in job satisfaction ( Humphrey et al., 2007 ), others reported that work context characteristics had an important role in job satisfaction and different indicators of performance ( Conlon & Dyne, 2004 ). Following this rationale we expected that social, knowledge, and work context characteristics would explain an important amount of variance for both job satisfaction and perceived performance. Hypothesis 4

Knowledge, social and work context characteristics will demonstrate incremental validity above the task characteristics for job satisfaction ( Hypothesis 4a) and perceived performance (Hypothesis 4b).

Method Translation

The translation of the WDQ into Spanish was accomplished through the translation/back-translation procedure recommended by Brislin (1980) . The researchers first translated the WDQ from English to Spanish and then a bilingual professional translator with experience in the Business Administration field back-translated the Spanish version into English. Following the translation from English to Spanish, we compared the original questionnaire to the back-translated English version and differences were resolved through discussion among authors; the professional translator was not aware of the study purpose. The wording of items was aimed to reflect general forms rather than specific idioms and expressions of the Spanish language as it is spoken in different countries. An initial version of the questionnaire was tested with a group of 18 workers from different occupational levels; once the questionnaire was administered, an interview with these workers was conducted in order to identify problems with the language expressions or wording. The resulting Spanish questionnaire used in the validation is presented in the Appendix.

Participants and Procedure

The sample was collected in Colombia, a country classified as an upper-middle-income economy with a GDP per capita (purchasing power parity) of US$ 9,125 (USA, US$ 43,063), an average economic growth from 2007 to 2011 of 4.5%, an annual employment growth of 3.5%, and 46.4% of its total work force being wage and salaried workers (USA, 93.2%); 77.2% of these salaried workers are distributed across four main economic sectors: wholesale and retail service (26.4%), health and social work (19.9%), agriculture, forestry, and fishing (17.5%), and manufacturing (13.4%); this same four economic sectors represent the 40.5% distribution of salaried workers in the USA ( International Labour Organization, 2013a, 2013b ).

Eight hundred forty-one Colombian employees participated in the study; however, 10 questionnaires were not usable due to participants not responding the WDQ section, leaving 831 useful questionnaires. The mean age for all participants was 34.9 years (range: 18-70, SD = 11 years); the mean tenure for all participants was 6.05 years (range: 1-51, SD = 7.48 years); 43% of workers were females (98.5 valid percent), 7.7% of respondents had completed education only at a high school/diploma level, 69.9% had completed undergraduate level (university, technical, or technological education), and 16.2% had completed postgraduate level or higher (93.8 valid percent).

Data were collected from three different samples. Sample 1 consisted of 279 full-time employees working for an organization that manufactures pumps, compressors, and valves (83.5% men; age: M = 31.2 years, SD = 8.2 years; education: M = 14.4 years, SD = 3.5 years); this sample d both blue and white collar workers. Sample 2 consisted of 89 full-time administrative employees working for a university (37% men; age: M = 39.3 years, SD = 8.8 years; education: M = 14.3 years, SD = 5.2 years). Sample 3 consisted of 463 full-time employees working for different organizations (44% men; age: M = 36.2 years, SD = 12.3 years; education: M = 16.2 years, SD = 4.1 years). Sample 3 was obtained in the context of an organizational behavior course with junior-level business administration students. These students d the work of a family member or acquaintance (job incumbent) who has worked full time for at least one year, and administrated the WDQ to the job incumbent (the students received specific training on the application of the questionnaire from one of the authors). This particular sampling strategy was employed so data could be collected on a wide range of jobs following the strategy used in the original work of Morgeson and Humphrey (2006) and is used when the goal is to sample a wide range of different jobs (e.g., Raymark, Schmit, & Guion, 1997 ). Workers in all samples filled a paper-and-pencil version of the WDQ and were informed about the confidential use of all information provided. The procedures were approved by the ethics review board of Pontificia Universidad Javeriana before the study began.

The distribution of the whole sample, depending on the economic sector in which job incumbents work, can be seen in Table 1 . This distribution s 17 out of 21 economic activities considered in the International Standard Industrial Classification of All Economic Activities (ISIC) (United Nations, 2008 ). However, it is important to note the under-representation of the agriculture, forestry and fishing sector, which has a very important role in the Colombian economy. The explanation for this under-representativeness is because sample was collected in an urban context and the access to these workers is difficult. Nevertheless, apart from this sector, the sample reflects the general formal labor market composition in Colombia.

Table 1

Incumbent Population by ISIC a

ISIC occupation category  n  Age (years) Job tenure (years) Sex(% men) 
    M  SD  M  SD   
Agriculture forestry and fishing  44.00  8.52  7.85  9.53  50 
Mining and quarrying  14  41.43  11.91  7.98  9.21  50 
Manufacturing  327  33.06  9.66  4.35  5.96  78 
Construction  42  34.13  11.15  4.49  8.84  60 
Wholesale and retail service  25  33.96  12.92  4.72  7.26  48 
Transportation and storage  43.80  11.30  17.12  21.13  80 
Accommodation and food service activities  29.78  13.34  3.43  4.70  22 
Information and communication  22  37.62  13.70  4.55  6.12  50 
Financial and insurance activities  37  36.19  11.49  8.16  7.61  49 
Real estate activities  41.40  17.44  4.82  4.53  60 
Professional, scientific and technical activities  29  33.55  11.30  3.52  4.65  24 
Administrative and support service activities  34.56  7.13  6.25  8.57  67 
Public administration and defense; compulsory social security  33  41.50  12.67  10.24  9.12  24 
Education  128  38.93  10.12  10.07  8.16  33 
Human health and social work activities  33  41.18  10.74  10.00  9.42  33 
Arts, entertainment and recreation  36.00  12.73  11.38  15.50  100 
Other services activities  35.36  7.86  2.00  1.03  20 
No information  102  29.70  10.92  4.48  5.30  47 
Total  831  34.92  11.12  6.05  7.47  56 

Note. All samples d.

a

International Standard Industrial Classification of All Economic Activities.

Occupational classification . We d five broad occupational categories (i.e., professional, non-professional, health, commercial, and manufacturing occupations) to test the occupational focused hypothesis ( H2a, H2b, H2c, H 2d). For professional and nonprofessional classification, job incumbents self-reported if their occupation was professional (their work activities a university degree) or nonprofessional (their work activities do not a university degree). For the other categories (health, commercial, and manufacturing occupations), job incumbents self-reported the economic sector in which their organization was located using the ISIC classification ( Table 1 ). For sample 1 (manufacture organization), all workers were classified in the manufacture sector; for sample 2 (university), all workers were classified in the education sector; for sample 3 (different organizations), business administration students helped the job incumbents to report this information. The “health” category was composed of the jobs within the “human health and social work activities” of the ISIC classification, whereas the “non-health” category was composed of the jobs in the remaining occupations. The “commercial” category was composed of the jobs within the “wholesale and retail service” of the ISIC classification. Finally, the “manufacturing” category was composed of the jobs within the “manufacturing” group in the ISIC classification.

A one-way ANOVA was conducted in order to verify any possible differences in demographic variables among the samples. For age, tenure, and education, significant differences were detected: age, F(2, 802) = 26.02, p ? .01 (sample 1, M = 31.2 years, sample 2, M = 39.3 years, sample 3, M = 36.2 years); tenure, F(2, 811) = 42.88, p ? .01 (sample 1, M = 3.1 years, sample 2, M = 10.3 years, sample 3, M = 7 years); education, F(2, 778) = 19.21, p ? .01 (sample 1, M = 14.4 years, sample 2, M = 14.3 years, sample 3, M = 16.2 years). Because of these results, analyses that use full-sample will be performed using these three variables as controls on the regression model of hypotheses 4a and 4b.

Measures

Work design. We used the WDQ developed by Morgeson and Humphrey (2006) that is a self-reporting measure that s 77 items using a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). Cronbach's alpha of the original version of the WDQ ranged from .64 (ergonomics) to .95 (task variety and physical demands), with a mean alpha = .86.

Job satisfaction was measured using the 4-item job satisfaction scale developed by De Witte (2000) in its Spanish version used by Isaksson (2007) . The scale uses a 5-point Likert scale ranging from 1 ( nothing important) to 5 (extremely important ). The Spanish version Cronbach's alpha = .81. An example item is: “La mayoría de los días estoy entusiasmado/a con mi trabajo.”

Perceived performance was measured through the 6-item perceived performance scale developed by Abramis (1994) in its Spanish version used by Isaksson (2007) . This scale asks respondents to think about their previous week at work and to rate how well they performed on six tasks, namely decision-making, performing without making mistakes, goal attainment, effort, taking initiatives, and taking responsibility. Five response categories were used, ranging from 1 ( very badly) to 5 (very well ). The Spanish version Cronbach's alpha = .79. An example item is: “¿En qué medida ha realizado satisfactoriamente la siguiente tarea? - Trabajar sin cometer errores.”

Analyses

To test hypothesis 1, we used confirmatory factor analysis (CFA) conducted with AMOS 21 ( Arbuckle, 2012 ). We applied a maximum likelihood estimation method; in order to run this analysis, no missing data will be allowed; taking into account that the missing data pattern was completely at random ( Schafer & Graham, 2002 ) and that the missing data percentage was less than 2% per variable, a mean imputation was done using SPSS. Maximum skewness was -1.74 and maximum kurtosis was 3.83, which is among the recommended bounds for skewness |2.0| and kurtosis |7.0| ( Hancock & Mueller, 2010).

Consistent with Morgeson and Humphrey (2006) , we report various fit indices: ? 2/df ratio, comparative fit index (CFI), standardized root-mean-square residual (SRMR), and root-mean-square error of approximation (RMSEA). For ? 2/df ratio, a value of 2.0 indicates good fit. For CFI, values higher than .90 indicate good fit. For SRMR, values lower than .08 indicate good fit. For RMSEA a value of .05 indicate good fit. We used the chi-square difference ( ??2 ) to compare the models, and accepted the more parsimonious model if it was not significantly different from a more complex model. We tested five different models of work design using CFA techniques:

  • 4-factor model examines the four broad categories of work characteristics (i.e., task, knowledge, social, and work context characteristics).

  • 18-factor model examines the work characteristics without any divisions (i.e., autonomy and interdependence as unique factors).

  • 19-factor model separates interdependence into its received and initiated components.

  • 20-factor model separates autonomy into its three components, which s autonomy in work scheduling, decision making, and work methods.

  • 21-factor model separates both interdependence and autonomy into the identified components.

To test hypotheses 2a, 2b, 2c and 2d we used t -tests to compare means of two occupational segments: professional and non-professional workers, health and manufacturing workers, commercial and manufacturing workers. To test hypothesis 3a and 3b we performed bivariate Pearson correlations; finally, to test hypothesis 4a and 4b, we used hierarchical regression analyses aiming to compute incremental validity, which determined the variance accounted for knowledge, social, and work context characteristics in job satisfaction and perceived performance beyond the variance explained by task characteristics.

Results

Table 2 presents the descriptive statistics of WDQ Scales. The first two columns present the means and standard deviations; overall there is no evidence of floor or ceiling effects. The third column presents Cronbach's alpha (?). In overall, the scales of the Spanish version of the WDQ demonstrate very good internal consistency reliability, and only in ergonomics the coefficients are somewhat low in ? (bellow .6). The fourth, fifth, and sixth columns present interrater reliability (intraclass correlation or ICC[2]; Bliese, 2000) and interrater agreement (r wg(j); James, Demaree, & Wolf, 1984 ). The ICC[2] assesses the extent to which incumbent judgments of their ISIC categories covary with each other relative to incumbents in other ISIC categories. The r wg(j) reflects the absolute level of agreement across raters and thus assesses the extent to which raters make similar mean-level ratings across their ISIC categories ( Morgeson & Humphrey, 2006 ). Generally, these statistics suggest that the incumbents within an ISIC occupational category agree on their work characteristics assessment. As in the original WDQ, there were some exceptions: task variety, task significance, job complexity, and problem solving. These variables demonstrate essentially zero interrater reliability. As Morgeson and Humphrey (2006) pointed, this could be due to a “lack of between-job variability in this sample or that perhaps these aspects of work are not stable characteristics of a job and reflect idiosyncratic elements of job holders” instead ( Morgeson & Humphrey, 2006 , p. 1326). Yet, the high levels of interrater agreement ( r wg(j) ) would suggest that these are not idiosyncratic perceptions because multiple incumbents agreed in their perceptions. Because job incumbents among an occupational category could hold different hierarchical levels, additional ICC[2] analyses using hierarchical level as the grouping variable were performed; results showed in Table 2 present ICC[2] values ranging from .54 to .97, p ? .05, for 20 out of 21 work characteristics. Taken as a whole, these data suggest that it is appropriate to aggregate to the occupational category (ISIC) and there are high levels of agreement about a job's category on work characteristics.

Table 2

Means, Standard Deviations and Reliability.

Construct  M  SD  Internal consistency a   ISIC Interrater reliability b   Hierarchical level Interrater reliability b   Interrater agreement c   Convergent-discriminant validity
              AVE  MSV 
Task characteristics
Work scheduling autonomy  3.87  0.88  .86  .64 **   .97 **   .80  .68  .68 
Decision-making autonomy  3.59  0.87  .84  .50 **   .97 **   .77  .65  .70 
Work methods autonomy  3.72  0.82  .83  .61 **   .96 **   .80  .61  .70 
Task variety  4.05  0.84  .91  ?.03  .87 **   .82  .71  .11 
Task Significance  4.20  0.65  .76  .18  .83 **   .88  .49  .12 
Task identity  3.90  0.72  .82  .70 **   .89 **   .88  .57  .30 
Feedback from job  4.00  0.75  .87  .54 **   .66 *   .88  .70  .30 
Knowledge characteristics
Job complexity  3.47  0.87  .80  ?.49  .93 **   .74  .58  .13 
Information processing  3.98  0.76  .80  .45 *   .97 **   .84  .51  .48 
Problem solving  3.63  0.74  .67  .07  .92 **   .75  .36  .48 
Skill variety  3.89  0.77  .91  .54 **   .95 **   .85  .73  .48 
Specialization  3.63  0.79  .85  .58 **   .95 **   .80  .59  .36 
Social characteristics
Social support  4.00  0.65  .81  .69 **   .54   .92  .44  .17 
Initiated interdependence  3.56  0.89  .75  .68 **   .56 *   .68  .53  .53 
Received interdependence  3.49  0.89  .75  .41 *   .77 **   .67  .52  .53 
Interaction outside organization  3.27  1.14  .92  .85 **   .94 **   .63  .75  .12 
Feedback from others  3.23  0.93  .84  .73 **   .96 *   .74  .65  .17 
Work context characteristics
Ergonomics  3.54  0.82  .57  .66 **   .92 **   .69  .49  .34 
Physical demands  2.31  1.10  .95  .79 **   .98 **   .73  .86  .21 
Work conditions  3.43  0.88  .76  .86 **   .97 **   .67  .40  .34 
Equipment use  2.85  0.96  .75  .65 **   .88 **   .50  .50  .21 

Note. All samples d. ? = Cronbach's alpha; AVE = average variance ed; MSV = maximum shared squared variance.

a

Coefficient alpha.

b

ICC[2].

c

r wg(j)

p < .10, * p < .05, ** p < .01.

Confirmatory Factor Analysis

In order to test the first hypothesis, that the Spanish WDQ version will fit a 21-factor structure, we run a set of five different CFAs. The results of our CFAs are presented in Table 3 , broken down for each sample. First, for the full sample, the 4-factor model showed poor fit, as the fit indexes were all off the accepted levels. Second, the 18-, 19-, and 20-factor solutions showed adequate fit, with the SRMR and RMSEA reaching adequate levels, whereas the CFI was slightly low and the ? 2 /df was slightly high. Finally, we tested the 21-factor solution; this model was the best fitted model overall, with the lowest ? 2/df ratio, SRMR, RMSEA and the highest CFI. This model was significantly better than the 18-factor model (? 2 change = 627, df change = 57, p < .001), the 19-factor model (? 2 change = 424, df change = 39, p < .001) and the 20-factor model (? 2 change = 204, df change = 20, p < .001). These same results patterns are present in the separate analyses for each sample ( Table 3 ); thus, like the original WDQ, the 21-factor model, which separates interdependence in two factors and autonomy in three factors, fits our data the best.

Table 3

Results of Confirmatory Factor Analysis.

Model  ? 2  df  ? 2/df ratio  SRMR  RMSEA  CFI 
4-factor
Full sample  21544  2843  7.58  .12  .09  .49 
Sample 1  8888  2843  3.13  .12  .08  .44 
Sample 2  7319  2843  2.57  .14  .13  .31 
Sample 3  14562  2843  5.12  .13  .09  .47 
18-factor
Full sample  6972  2696  2.59  .06  .04  .88 
Sample 1  4659  2696  1.73  .07  .05  .82 
Sample 2  5503  2696  2.04  .11  .11  .57 
Sample 3  5613  2696  2.08  .06  .05  .87 
19-factor
Full sample  6769  2678  2.53  .06  .04  .89 
Sample 1  4606  2678  1.72  .07  .05  .82 
Sample 2  5449  2678  2.04  .11  .11  .58 
Sample 3  5468  2678  2.04  .06  .05  .87 
20-factor
Full sample  6549  2659  2.46  .06  .04  .89 
Sample 1  4523  2659  1.70  .07  .05  .83 
Sample 2  5329  2659  2.00  .11  .11  .59 
Sample 3  5342  2659  2.01  .06  .05  .88 
21-factor
Full sample  6345  2639  2.40  .06  .04  .90 
Sample 1  4470  2639  1.69  .07  .05  .83 
Sample 2  5274  2639  2.00  .11  .11  .60 
Sample 3  5197  2639  2.00  .06  .05  .89 

Note. SRMR = standardized root-mean-square; RMSEA = root-mean-square error of approximation; CFI = comparative fit index.

Additional evidence for a 21-factor structure is provided when the structure of task and social characteristics are studied separately. For task characteristics, a comparison between 5- and 7-factor model was developed (separating or integrating the three dimensions of autonomy). The 7-factor model performed better in ? 2 /df ratio in comparison to the 5-factor model (? 2 change = 372, df change = 11, p < .001), SRMR = .08, RMSEA = .07, CFI = .93). For social characteristics, a comparison between a 4- and a 5-factor model was developed (separating or integrating the two dimensions of interdependence); the 5-factor model performed better in ? 2 /df ratio than the 4-factor model (? 2 change = 184, df change = 4, p < .001), SRMR = .06, RMSEA = .07, CFI = .92). In sum, these data fully support hypothesis 1.

Convergent/Discriminant Validity

Once we confirmed the 21-factor measurement model, it is important to assess the extent to which the items of a specific factor converge or share a high proportion of variance (convergent validity). In addition, it is also important to assess the extent to which a factor is truly distinct from other factors both in terms of how much it correlates with other factors and how distinctly items represent only this single factor (discriminant validity).

To evaluate convergent validity, two methods were used: (a) assessment of standardized factor loadings of observable variables and (b) average variance ed (AVE) for each factor. For the first method, a comparison of factor loadings of each item was conducted, loadings estimates should be significant and with a factor loading of .50 or higher for the associated item; for the AVE method, “the average variance ed is calculated as the mean variance ed for the item loading on a [factor] and is a summary indicator of convergence” ( Hair, Black, Babin, & Anderson, 2010 , p. 687) and values greater than .50 are considered adequate.

On the other hand, to evaluate discriminant validity, a comparison between the AVE and the maximum shared squared variance (MSV) of each factor was carried out; the MSV represents the maximum shared squared variance found when comparing for any two factors (in this case, each factor was compared with all other 20 factors); with the square of the correlation estimate between these two factors, the AVE estimates should be greater than the MSV; this is because “a latent [factor] should explain more of the variance in its item measures that it shares with another construct” ( Hair et al., 2010 , p. 688); when the MSV values are lower than the AVE, the measure has good discriminant validity.

The results for convergent validity indicate that the standardized factor loadings of the WDQ indicate average estimate of .74, with just 4 out of 77 items below .50 (items 15, 18, 25, and 66). 1 1

Full standardized factor loadings are available from the first author.

For the AVE method, as shown in Table 2 , the values of 5 out of 21 factors were below .50: task significance .49, problem solving .36, social support .44, ergonomics .49, and work conditions .40. Taken together, the evidence supports the convergent validity of the measurement model (with some limitations in the factors just mentioned).

On the other side, for discriminant validity, some values of MSV were below the AVE values, indicating some issues in the autonomy factors (work scheduling, decision making, and work methods), interdependence factors (initiated and received) and problem solving factor; however, the other 15 factors showed adequate levels of divergent validity.

Differences between Occupations

As in the original WDQ validation, our second set of hypotheses suggested that jobs within broad occupational categories would differ in certain work characteristics. First, hypothesis 2a predicted that knowledge and autonomy characteristics would be higher for jobs in professional than in nonprofessional occupations. As shown in Table 4 , jobs in professional occupations had higher levels for all knowledge characteristics: job complexity, t(677) = 6.56, r 2 = .05, p < .001; information processing, t(789) = 9.79, r 2 = .11, p < .001; problem solving, t(654) = 5.98, r 2 = .05, p < .001; skill variety, t(789) = 6.61, r 2 = .05, p < .001; and specialization t(789) = 3.93, r 2 = .02, p < .001. Also, all autonomy characteristics were higher in professional occupations: work scheduling autonomy, t(789) = 10.96, r 2 = .14, p < .001; decision-making autonomy, t(696) = 9.54, r 2 = .11, p < .001; and work methods autonomy, t(789) = 8.90, r 2 = .09, p < .001. Thus hypothesis 2a was fully supported.

Table 4

Means of Jobs across Occupational Categories.

Work characteristics  Occupational category
  Professional  Nonprofessional 
Job complexity , **   3.63  3.23 
Information processing **   4.20  3.69 
Problem solving , **   3.76  3.44 
Skill variety **   4.03  3.68 
Specialization **   3.71  3.48 
Work scheduling autonomy **   4.15  3.51 
Decision-making autonomy , **   3.82  3.26 
Work methods autonomy **   3.93  3.43 
Physical demands **   1.91  2.79 
Work conditions **   3.67  3.14 
  Health-focused  Manufacturing 
Task significance , *   4.52  4.24 
  Commercial  Manufacturing 
Interaction outside organization , **   3.86  2.77 

Note. All samples d.

Equal variances not assumed.

*

p < .05,

**

p < .001.

Second, hypothesis 2b predicted that jobs in nonprofessional occupations would have higher levels of physical demands and lower levels of work conditions than jobs in professional occupations. As shown in Table 4 , this hypothesis was supported, as jobs in nonprofessional occupations had higher levels for physical demands, t(789) = 12.20, r 2 = .16, p < .001, and lower levels of work conditions, t(789) = 8.84, r 2 = .09, p < .001. Third, hypothesis 2c predicted that jobs in health-focused occupations would have higher levels of task significance than manufacturing jobs. As shown in Table 4 , this hypothesis was also supported, as the jobs in the health-focused occupations had higher levels of task significance, t(41) = 2.78, r 2 = .06, p < .05. Finally, hypothesis 2d predicted that jobs in commercial occupations would have higher levels of interaction outside organization than manufacturing occupations. This hypothesis was also supported as interaction outside organization was higher for jobs in commercial occupations than in manufacturing occupations, t(29) = 4.97, r 2 = .21, p < .001.

Relationshi Work Characteristics and Outcomes

Hypothesis 3a predicted that task, knowledge, and social characteristics would be positively related to job satisfaction. As shown in Table 5 , all seven task characteristics were significantly related to job satisfaction, ranging in magnitude from .18 to .34 (mean correlation of .29). On the other hand, 4 out of 5 knowledge characteristics were significantly related to job satisfaction, ranging in magnitude from .15 to .23 (mean correlation of .19). Finally, for the social characteristics, only social support (.30) and feedback from others (.12) were related to job satisfaction, thus hypothesis 3a was supported for 13 out of 17 work characteristics.

Table 5

Intercorrelations among Study Variables.

  10  11  12  13  14  15  16  17  18  19  20  21  22   
Task characteristics
1. Work scheduling autonomy  –                                             
2. Decision-making autonomy  .69 **   –                                           
3. Work methods autonomy  .69 **   .72 **   –                                         
4. Task variety  .27 **   .31 **   .27 **   –                                       
5. Task significance  .24 **   .27 **   .25 **   .35 **   –                                     
6. Task identity  .37 **   .33 **   .36 **   .22 **   .33 **   –                                   
7. Feedback from job  .29 **   .33 **   .37 **   .27 **   .37 **   .57 **   –                                 
Knowledge characteristics
8. Job complexity  .06  .08 *   .06  .23 **   .09 *   ?.09 *   ?.05  –                               
9. Information processing  .28 **   .29 **   .32 **   .43 **   .33 **   .23 **   .26 **   .35 **   –                             
10. Problem solving  .22 **   .31 **   .29 **   .31 **   .30 **   .09 *   .19 **   .17 **   .52 **   –                           
11. Skill variety  .23 **   .30 **   .29 **   .42 **   .38 **   .21 **   .24 **   .26 **   .55 **   .56 **   –                         
12. Specialization  .18 **   .29 **   .27 **   .21 **   .36 **   .25 **   .26 **   .13 **   .42 **   .41 **   .55 **   –                       
Social characteristics
13. Social support  .24 **   .26 **   .25 **   .22 **   .35 **   .36 **   .38 **   ?.06  .18 **   .16 **   .21 **   .21 **   –                     
14. Initiated interdependence  .10 **   .13 **   .15 **   .24 **   .25 **   .10 **   .10 **   .10 **   .27 **   .24 **   .26 **   .24 **   .13 **   –                   
15. Received interdependence  .08 *   .10 **   .12 **   .21 **   .20 **   .05  .07  .13 **   .29 **   .19 **   .23 **   .18 **   .07 *   .56 **   –                 
16. Interaction outside organization  .14 **   .17 **   .13 **   .16 **   .21 **   .20 **   .18 **   .03  .20 **   .17 **   .14 **   .08 *   .30 **   .09 **   .17 **   –               
17. Feedback from others  .10 **   .13 **   .09 *   .11 **   .14 **   .22 **   .29 **   ?.07 *   .14 **   .09 *   .06  .13 **   .43 **   .07 *   .04  .27 **   –             
Work context characteristics
18. Ergonomics  .27 **   .24 **   .25 **   .11 **   .12 **   .27 **   .20 **   ?.01  .13 **   .02  .07 *   .08 *   .30 **   .06  .07 *   .09 **   .19 **   –           
19. Physical demands  ?.24 **   ?.16 **   ?.17 **   ?.03  .01  ?.12 **   ?.07 *   ?.15 **   ?.15 **   ?.02  ?.01  .10 **   ?.07 *   .08 *   ?.03  ?.08 *   ?.00  ?.45 **   –         
20. Work conditions  .29 **   .20 **   .25 **   .08 *   .06  .25 **   .19 **   ?.05  .15 **   .04  .05  .02  .24 **   .01  .04  .17 **   .19 **   .46 **   ?.40 **   –       
21. Equipment use  ?.05  .04  .02  .10 **   .18 **   .06  .11 **   ?.06  .10 **   .11 **   .19 **   .37 **   .06  .20 **   .16 **   .03  .11 **   ?.06  .41 **   ?.11 **   –     
Outcomes
22. Job satisfaction  .34 **   .33 **   .31 **   .18 **   .27 **   .31 **   .31 **   .05  .15 **   .15 **   .23 **   .23 **   .30 **   .03  .04  .06  .12 **   .25 **   ?.11 **   .17 **   .07  –   
23. Perceived performance  .28 **   .32 **   .33 **   .26 **   .35 **   .30 **   .35 **   .08 **   .29 **   .26 **   .37 **   .40 **   .31 **   .20 **   .15 **   .04  .13 **   .17 **   .02  .09 **   .17 **   .37 **    

Note. All samples d. n ranges from 810 to 831.

*

p < .05,

**

p < .01.

Hypothesis 3b predicted that task, knowledge, and social characteristics would be positively related to perceived performance. As shown in Table 5 , all seven task characteristics were significantly related to perceived performance, ranging in magnitude from .26 to .35 (mean correlation of .31). Also all five knowledge characteristics were significantly related to perceived performance, ranging in magnitude from .08 to .40 (mean correlation of .28). Finally, four out of five social characteristics were significantly related to perceived performance, ranging in magnitude from .13 to .31 (mean correlation of .19). Thus, hypothesis 3b was supported for 16 out of 17 work characteristics.

The last set of hypotheses predicted that knowledge, social, and context characteristics would incrementally predict job satisfaction (4a) and perceived performance (4b) beyond task characteristics. To test these hypotheses, we conducted a hierarchical regression in which we first regressed job satisfaction or perceived performance on three control variables (age, organizational tenure, and education) as the first step, task characteristics as the second step, knowledge characteristics as the third step, social characteristics as the fourth step, and work context characteristics as the fifth step. As shown in Table 6 , for the job satisfaction model, control variables explained only small amounts of variance (5%); however, when knowledge, social, and work context characteristics are introduced, they explained additional amounts of variance ? R 2 = .05, p < .01. On the other hand, for the perceived performance model, the introduction of control variables explained small amounts of variance (6%) and the introduction of knowledge, social, and work context characteristics explained additional amounts of variance, ? R 2 = .08, p < .01.

Table 6

Incremental Validity of Work Characteristics on Job Satisfaction and Perceived Performance.

  Work outcomes
  Job satisfaction Perceived performance
Predictor  Sample 1  Sample 2  Sample 3  All samples  Sample 1  Sample 2  Sample 3  All samples 
Step 1 R 2 Age, Tenure, Education  .06 **   .06  .05 **   .05 **   .07 **   .05  .09 **   .06 **  
Step 2 ?R 2Task Characteristics  .19 **   .18 *   .18 **   .16 **   .17 **   .31 **   .25 **   .20 **  
Step 3 ?R 2 Knowledge Characteristics  .02 **   .05 *   .02 **   .01 **   .07 **   .17 **   .01 **   .05 **  
Step 4 ?R 2Social Characteristics  .03 **   .15 **   .04 **   .02 **   .04 **   .05 **   .03 **   .02 **  
Step 5 ?R 2 Work Context Characteristics  .03 **   .01 *   .02 **   .02 **   .01 **   .03 **   .00 **   .01 **  
Total R 2  .33 **   .45 *   .31 **   .26 **   .36 **   .61 **   .38 **   .34 **  
**

p < .01,

*

p < .05.

Taking these small values into account, additional analyses were conducted for each sample. As shown in Table 6 , the social characteristics explain medium to small amounts of variance for job satisfaction in the university, ? R 2 = .15, p < .01, manufacture, ?R 2 = .03, p < .01, and different organization sample ? R 2 = .04, p < .01. On the other hand, knowledge characteristics explained significant amounts of variance for perceived performance in the university, ? R 2 = .17, p < .01, and manufacture organization ? R 2 = .07, p < .01 samples. Therefore, our results (general and segregated) provide some support for hypothesis 4a and 4b, especially for knowledge and social characteristics; however, work context characteristics explained only a small amount of variance for both job satisfaction and perceived performance.

Common Method Bias

As with all self-reported data, there is a potential for the occurrence of common method bias; in order to control this, we used the common latent factor method (CLF), in which “items are allowed to load on their theoretical constructs, as well as on a latent common methods variance factor, and the significance of the structural parameters is examined both with and without the latent common methods variance factor in the model” ( Podsakoff, MacKenzie, Lee, & Podsakoff, 2003 , p. 891). Taking this into account, we compared the original 21-factor model for the full sample ( Table 3 ) and a 21-factor model in which the 77 items were allowed to load on their original factors, as well as on a CLF. The results for the 21-factor CLF model (? 2 = 6125, df = 2635, p < .001, ?2 /df = 2.33, CFI = .90, SRMR = .06, RMSEA = .04) produced a small change in the model fit (? 2 change = 219, df change = 4, p < .001, ?2 /df change = .07, no changes in CFI, SRMR, or RMSEA), what represented an slightly improved fit compared with the original 21-factor model presented in Table 3 ; however, when analyzing the significance of the structural parameters in both models (with and without the CLF), no significant changes in parameter estimates were found between the two models, these results indicating that the amount of variance due to common method bias is relatively small.

Discussion

The purpose of this study was to test the validity of a Spanish version of the WDQ developed by Morgeson and Humphrey (2006) . The questionnaire was administrated to 831 job incumbents working in 17 ISIC economic sectors. The CFA results indicated support for a 21-factor solution; this is in line with previous validations in German ( Stegmann et al., 2010) and Italian (Zaniboni et al., 2013 ). Furthermore, the internal consistency reliabilities for almost all scales are above .70. As in the original validation ( Morgeson & Humphrey, 2006 ), the Spanish version of the WDQ was able to detect expected differences in work characteristics across different sets of occupations, providing construct validity evidence. In addition, we found that knowledge, social, and context characteristics incrementally predict job satisfaction and perceived performance beyond task characteristics.

As our results supported the validity of the Spanish version of the WDQ, the study further contributes to the generalization of work characteristics taxonomy proposed by Morgeson and Humphrey (2006) . Thus, our results from three Colombian samples provide further evidence for the generalizability of the scale in different cultural settings. The WDQ validation in a cultural setting different from the one in which they were developed and validated (USA, Western Europe) gives additional support to the structure of work characteristics that are relevant for all works and organizations.

The confirmation of the 21-factor structure was expected, as Colombia during the last 20 years has opened its borders to new organizations and work arrangements and has begun to switch from an agriculture and production economy to a more intense service economy ( Ogliastri, 2007 ). This is consistent with the importance given to social and knowledge characteristics which are distinctive of service organizations beyond the traditional task and work context characteristics that are typical of production industries.

The Spanish version of the WDQ obtained better psychometric results than similar work characteristics instruments that were tested in the Spanish speaking countries context such as the JDS ( Fuertes et al., 1996 ) and presented a clear internal structure of the model, showing high reliability among almost all 21 work characteristics. The work context characteristic that showed some low reliability (ergonomics) d a reverse coded item, a relatively common phenomena that is also presented in other organizational behavior questionnaire validations ( Podsakoff et al., 2003 ). However, when the problem item is eliminated, the factor reliability improves considerably (? = .84). In addition, this work characteristic was the one with the lowest reliability in the original validation of Morgeson and Humphrey (2006).

Although all hypotheses were supported in some degree, two unexpected relations emerged in the present Spanish validation. The first unexpected result was the differential role of knowledge and social characteristics in job satisfaction and perceived performance ( Table 6). The ?R 2 in job satisfaction (14%) due to social characteristics, and in perceived performance (19%) due to knowledge characteristics in the university sample may be related with the observations of Michael Campion in the interdisciplinary approach to work design ( Campion, 1988 ), that states that different job design approaches influence different outcomes (e.g., the motivational approach is more correlated with satisfaction outcomes than the rest of the approaches); in this case, what was found was that in this sample, some work characteristics influence different outcomes. These results could indicate that when a work redesign is imminent (especially in knowledge-oriented organizations as the university), it is important to pay attention to which specific outcomes are of interest to change and depending on this, it will be necessary to evaluate only some work characteristics.

The second unexpected result was related to some differences in the convergent and discriminant validity results on autonomy and social characteristics ( Table 2 ). From the psychometric perspective, these results on low divergent validity among autonomy (work scheduling, decision-making and, work methods) and interdependence (initiated and received) may be explained by the fact that they are composites of a more general factor (autonomy and interdependence) and for this reason it is expected that MSV values will be greater than those in the other variables.

Finally, we have to be cautious with the problem solving variable, due to its non-significant interrater reliability, low AVE, and high MSV. These results indicate that this particular variable share a high portion of variance with other variables, in particular with information processing. Because of this, the analysis of these variables should be treated with caution in future Spanish WDQ administrations.

Implications for Practice and Research

For practitioners, a broader range of work design potentials is possible beyond the traditional five job characteristics of the JDS; however, it is important that practitioners be fully aware of potential cultural influences in the work design practice. In the case of Latin-America countries, a deep collectivism value can be found, and in Colombia there is an important role of collectivism and rejection to individualism ( Ogliastri, 2007 ). In an article from the GLOBE project ( Dorfman, Javidan, Hanges, Dastmalchian, & House, 2012 ), the authors report that the autonomous dimension (tendencies to act independently without relying on others) is strongly negative related to institutional collectivism (degree in which organizational practices encourage and reward collective distribution of resources and collective action); this relation is important in order to redesign (increase) autonomy in positions where high institutional collectivism is present.

With regard to research, the WDQ is a tool that allows to investigate the impact of different work configurations on organizational and personal outcomes and let open a research line of the influence of different mediators and moderators in the relation between work characteristics and personal and organizational outcomes (e.g., cultural characteristics).

Limitations

Two major drawback limited the present research: first, the presence of some level of common method bias, which implies that the results must be interpreted with caution, even though the CLF test results indicated that variance due to common method is between the acceptable limits. This is consistent with previous research on common method variance, which has concluded that while common method bias may be present, it may not always significantly affect the results and conclusions drawn from the data ( Crampton & Wagner, 1994; Doty & Glick, 1998 ). The second limitation was the sample method ion; although we used three different samples and almost all occupational groups of the ISIC were considered, some groups were sub-represented (e.g., arts, entertainment, and recreation); besides, it is also important to consider that half of the labor market in Colombia is informal and the conclusions of this study can apply only to the workers that are inside the formal labor market (50% of the Colombian total labor force). Future research should consider the structure of work characteristics in jobs d in the informal labour market. They represent in developing countries an important amount of the total economy and deserve a better analysis and understanding.

Conclusion

The Spanish version of the WDQ is a validated and reliable instrument to assess work characteristics in the Spanish speaking context. Our study provided evidence for the validity of a Spanish version of the scale and presented further support for the generalization of the 21-factor of work design characteritics in different cultural settings that d particular relations between knowledge and social characteristics and job satisfaction and perceived performance. We hope that the introduction of this instrument will stimulate further research and practice on work design in Spanish speaking countries.

Conflict of Interest

The authors of this article declare no conflict of interest.

Financial Support

This research has been founded by Pontificia Universidad Javeriana and sponsored by the project PSI2012-36557 funded by DGICYT and the funding of the Generalitat Valenciana for research groups of excellence PROMETEO 2012/048.

Referencias

References
Abramis, 1994
D.J. Abramis
Relationship of Job Stressors to Job Performance: Linear or Inverted-U?
Psychological Reports
75
1994
547-58
Arbuckle, 2012
J.L. Arbuckle
IBM SPSS AMOS (Version 21.0) [Computer Program]
2012
Bliese, 2000
P. Bliese
Within-group agreement, non-independence, and reliability: Implications for data aggregation and analysis
Multilevel theory, research, and methods in organizations
Jossey-Bass
San Francisco
2000
349-81
Boada et al., 2004
J. Boada
R. Vallejo
D. de
E. Agulló
El burnout y las manifestaciones psicosomáticas como consecuentes del clima organizacional y de la motivación laboral [Burnout and psychosomatic manifestations as consequence of organizational climate and work motivation]
Psicothema
16
2004
125-31
Brislin, 1980
R.W. Brislin
Translation and content analysis of oral and written material
Allyn and Bacon
H.C.TriandisJ.W.Berry
Boston, MA
1980
389-444
Campion, 1985
M.A. Campion
The Multimethod Job Design Questionnaire (MJDQ)
Psychological Documents
15
1985
221-49
Campion, 1988
M.A. Campion
Interdisciplinary approaches to job design: A constructive replication with extensions
Journal of Applied Psychology
73
1988
467-81
Campion and Thayer, 1985
M.A. Campion
P.W. Thayer
Development and field evaluation of an interdisciplinary measure of job design
Journal of Applied Psychology
70
1985
29-43
Conlon and Dyne, 2004
D. Conlon
L. Van Dyne
The effects of physical and social context on evaluations of captive, intensive service relationships
Academy of Management Journal
47
2004
433-45
Crampton and Wagner, 1994
S. Crampton
J. Wagner
Percept-percept inflation in microorganizational research: An investigation of prevalence and effect
Journal of Applied Psychology
79
1994
67-76
Dávila and Chacón, 2003
M. Dávila
F. Chacón
Adaptación de instrumentos para la evaluación de aspectos organizacionales en ONG's [Adapting instruments for assessing organizational aspects in NGO's]
Revista de Psicología del Trabajo y de las Organizaciones
19
2003
159-79
De Witte, 2000
H. De Witte
Arbeidsethos en Jobonzekerheid: Meting en Gevolgen voor Welzijn, Tevredenheid en Inzet op het Werk [Work Ethic and Job Insecurity: Assessment and Consequences for Well-being, Satisfaction and Performance at Work]
Van groep naar gemeenschap [From group to community]
Garant
Leuven
2000
325-50
Dorfman et al., 2012
P. Dorfman
M. Javidan
P. Hanges
A. Dastmalchian
R. House
GLOBE: A twenty year journey into the intriguing world of culture and leadership
Journal of World Business
47
2012
504-18
Doty and Glick, 1998
D.H. Doty
W.H. Glick
Common Methods Bias: Does Common Methods Variance Really Bias Results?
Organizational Research Methods
1
1998
374-406
Edwards et al., 1999
J. Edwards
J. Scully
M. Brtek
The measurement of work: Hierachical representation of the Multimethod Job Design Questionnaire
Personnel Psychology
52
1999
305-34
Edwards et al., 2000
J. Edwards
J. Scully
M. Brtek
The nature and outcomes of work: a replication and extension of interdisciplinary work-design research
The Journal of Applied Psychology
85
2000
860-8
Fried and Ferris, 1987
Y. Fried
G.R. Ferris
The Validity of the Job Characteristics Model: a Review and Meta-Analysis
Personnel Psychology
40
1987
287-322
Fuertes et al., 1996
Fuertes, F., Munduate, L., & Fortea, M. A. (1996). Análisis y rediseño de puestos (Adaptación española del cuestionario Job Diagnostic Survey - JDS) [Work analysis and redesign - Spanish adaptation of the Job Diagnostic Survey - JDS]. Castellón: Centro de publicaciones de la Universistat Jaume I.
Grant et al., 2010
A.M. Grant
Y. Fried
T. Juillerat
Work matters: Job design in classic and contemporary perspectives
APA handbook of industrial and organizational psychology, Vol 1: Building and developing the organization
American Psychological Association
Washington, DC, US
2010
417-53
Grant and Parker, 2009
A.M. Grant
S.K. Parker
Redesigning Work Design Theories: The Rise of Relational and Proactive Perspectives
Academy of Management Annals
3
2009
317-75
Hackman and Lawler, 1971
J.R. Hackman
E.E. Lawler
Employee reactions to job characteristics
Journal of Applied Psychology
55
1971
259-86
Hackman and Oldham, 1975
J.R. Hackman
G. Oldham
Development of the Job Diagnostic Survey
Journal of Applied Psychology
60
1975
159-70
Hackman and Oldham, 1976
J.R. Hackman
G.R. Oldham
Motivation through the design of work: test of a theory
Organizational Behavior and Human Performance
16
1976
250-79
Hackman and Oldham, 1980
J.R. Hackman
G.R. Oldham
Work Redesign
1980
Hair et al., 2010
J.F. Hair
W.C. Black
B.J. Babin
R.E. Anderson
Multivariate Data Analysis
7th ed.
2010
Hancock and Mueller, 2010
G. Hancock
R. Mueller
The reviewer's guide to quantitative methods in the social sciences
New York, NY: Routledge
2010
Hofstede, 2001
G. Hofstede
Culture's Consequences: Comparing Values, Behaviors, Institutions and Organizations Across Nations
2001
House et al., 2004
R.J. House
P.J. Hanges
M. Javidan
P.W. Dorfman
V. Gupta
Culture, Leadership, and Organizations: The GLOBE Study of 62 Societies
2004
Humphrey et al., 2007
S.E. Humphrey
J.D. Nahrgang
F.P. Morgeson
Integrating motivational, social, and contextual work design features: a meta-analytic summary and theoretical extension of the work design literature
The Journal of Applied Psychology
92
2007
1332-56
International Labour Organization, 2013a
International Labour Organization (2013a). Key Indicators of the Labour Market. (8th Edition).
International Labour Organization, 2013b
International Labour Organization (2013b). World of Work Report 2013 Repairing the economic and social fabric . Geneva, Switzerland.
Isaksson, 2007
Isaksson, K. (2007). Psychological Contracts across Employment Situations (PSYCONES) . Brussels. Retrieved from h t t p & # 5 8 ; & # 4 7 ; & # 4 7 ; w w w & # 4 6 ; u n i - l e i p z i g & #46;de/?apsycho/rigotti/Psycones_finalreport.pdf.
James et al., 1984
L.R. James
R.G. Demaree
G. Wolf
Estimating within-group interrater reliability with and without response bias
Journal of Applied Psychology
69
1984
85-98
Johns et al., 1992
G. Johns
J. Xie
Y. Fang
Mediating and moderating effects in job design
Journal of Management
18
1992
657-76
Kulik et al., 1988
C.T. Kulik
G.R. Oldham
P.H. Langner
Measurement of job characteristics: Comparison of the original and the revised Job Diagnostic Survey
Journal of Applied Psychology
73
1988
462-6
Martínez-Gómez and Marín-García, 2009
M. Martínez-Gómez
J.A. Marín-García
Cómo Medir y Guiar Cambios hacia Entornos Educativos más Motivadores [How to measure and lead changes to more motivating educational environments]
Formación Universitaria
2
4
2009
3-14
Morgeson and Humphrey, 2006
F.P. Morgeson
S.E. Humphrey
The Work Design Questionnaire (WDQ): developing and validating a comprehensive measure for assessing job design and the nature of work
Journal of Applied Psychology
91
2006
1321-39
Ogliastri, 2007
E. Ogliastri
Colombia: The Human Relations Side of Enterprise
Culture and Leadership Across the Word. The GLOBE Book of In-Depth Studies of 25 Societies
Psychology Press
New York
2007
689-722
Osca and Urien, 2001
A. Osca
B. Urien
Rediseño de tareas, satisfacción laboral y rendimiento: un estudio en la industria de la automoción [Task redesign, job satisfaction and performance: a study in the automotive industry]
Revista de Psicología del Trabajo y de las Organizaciones
17
2001
327-40
Podsakoff et al., 2003
P.M. Podsakoff
S.B. MacKenzie
J.-Y. Lee
N.P. Podsakoff
Common method biases in behavioral research: A critical review of the literature and recommended remedies
Journal of Applied Psychology
88
2003
879-903
Ramis et al., 2007
C. Ramis
M.A. Manassero
V. Ferrer
E. García-Buades
¡No es fácil ser un buen jefe/a! Influencia de las habilidades comunicativas de la dirección sobre la motivación, la autoeficacia y la satisfacción de sus equipos [It's not easy being a good boss: Influence of leadership communicative skills on motivation, self-efficacy, and satisfaction of their teams]
Revista de Psicología del Trabajo y de las Organizaciones
23
2007
161-81
Raymark et al., 1997
P.H. Raymark
M.J. Schmit
R.M. Guion
Identifying Potentially Useful Personality Constructs for Employee Selection
Personnel Psychology
50
1997
723-36
Roberts and Glick, 1981
K.H. Roberts
W. Glick
The job characteristics approach to task design: A critical review
Journal of Applied Psychology
66
1981
193-217
Salas and Glickman, 1990
E. Salas
A.S. Glickman
Comportamiento organizacional, teoría de sistemas socio-técnicos y calidad de vida laboral: la experiencia Peruana [Organizational behavior, socio-technical systems theory and quality of working life: the Peruvian experience]
Revista Latinoamericana de Psicología
22
1990
69-82
Schafer and Graham, 2002
J.L. Schafer
J.W. Graham
Missing data: Our view of the state of the art
Psychological Methods
7
2
2002
147-77
Solf, 2006
A. Solf
Motivación intrínseca laboral y su relación con las variables de personalidad orientación a la meta y tesón [Work intrinsic motivation and its relationship with personality variables goal orientation and conscientiousness]
Persona
9
2006
111-26
Stegmann et al., 2010
S. Stegmann
R. Van Dick
J. Ullrich
J. Charalambous
B. Menzel
N. Egold
Der Work Design Questionnaire Vorstellung und erste Validierung einer deutschen Version [The Work Design Questionnaire: Presentation and initial validation of a German version]
Zeitschrift Für Arbeits- U. Organisationspsychologie
54
2010
1-28
Taber and Taylor, 1990
T. Taber
E. Taylor
A review and evaluation of the psychometric properties of the Job Diagnostic Survey
Personnel Psychology
43
1990
467-500
Turner and Lawrence, 1965
A.N. Turner
P.R. Lawrence
Industrial Jobs and the Worker. An Investigation of Response to Task Attributes
1965
United Nations, 2008
United Nations (2008). International Standard Industrial Classification of All Economic Activities (ISIC) (Rev. 4). New York. Retrieved from h t t p & # 5 8 ; & # 4 7 ; & # 4 7 ; u n s t a t s & # 4 6 ; u n & # 46;org/unsd/publication/seriesM/seriesm_4rev4e.pdf.
Zaniboni et al., 2013
Zaniboni, S., Truxillo, D. M., & Fraccaroli, F. (2013, May). Validation of the Italian Version of the Work Design Questionnaire (WDQ). In S. Stegmann (Chair), SIOP-IAAP-EAWOP Alliance Symposium: Comprehensive Work Design Analysis - Insights from Around the Globe . Symposium conducted at the 16th EAWOP Congress, Münster, Germany.

Copyright © 2018. Colegio Oficial de Psicólogos de Madrid

© Copyright 2018. Colegio Oficial de Psicólogos de Madrid ContactPrivacy PolicyCookies Policy

We use our own and third­party cookies. The data we compile is analysed to improve the website and to offer more personalized services. By continuing to browse, you are agreeing to our use of cookies. For more information, see our cookies policy

Aceptar
h**a**