2.1. Conceptualization of high-performance practices

HPWPs originated in the 1990s as part of the Strategic Human Resources Management trend that sought to align management functions and activities in this area with the organization’s strategy (^{Kaushik and Mukherjee, 2021}). In the literature, HPWPs can be identified as high-performance work systems, high involvement work systems and high commitment management, among others (^{Raineri, 2017}). However, and regardless of the term used, among these denominations there is a common aspect related to obtaining a better performance of the company when they are used together forming a system and promoting through them, the development of skills and capabilities in employees (Kaushik and Mukherjee, 2021).

The definition of HPWP has been offered by various authors. For example, ^{Evans and Davis (2015)} conceptualize them as an integrated system of human resources practices that are internally consistent (alignment between the same practices) and externally consistent (alignment with organizational strategy). In the same line and under a more recent contribution, ^{Haar, O`Kane, and Daellenbach (2021)} define them as those practices capable of developing a superior workforce that contributes to enhance knowledge, skills and capabilities in employees, making a difference in the market competitors.

The interest aroused by this set of practices in the human resources area is originated by a growing number of researchers who agree with the positive effects of including HPWPs to improve company results (^{Arthur, 1994}; ^{Delaney and Huselid, 1996}; ^{Kloutsiniotis, Katou, and Mihail. 2021}; ^{Haar et al. 2021}; Sun and ^{Mamman, 2021}). Thus, different research has been conducted to determine the impact of HPWPs on variables such as performance (^{Raineri, 2017}), innovation (^{Haar et al. 2021}) and affective commitment (^{Para-Gonzalez, Jiménez-Jiménez, and Martínez-Lorente, 2019}), in addi­tion to recognizing that there is a crucial distinction between the intention with which HPWPs are carried out by managers and the perception that employees attribute to the implementation (^{Ma, Gong, Long, and Zhang, 2021}).

2.2. Classification and measurement of high-performance practices

Several proposals have been put forward by authors to classify HPWPs. However, there remains the absence of a commonly accepted classification, which would allow comparative studies based on the results obtained (^{Miao et al. 2021}). In this sense, a review of representative proposals in the literature that facilitate the operationalization of the concept is presented below.

For ^{Arthur (1994)}, HPWPs are classified into two dimensions called control and commitment, which could shape the attitudes and behaviors of individuals at work. ^{Huselid (1995)} argues that the HPWPs adopted by an organization should be aligned with the aspects that are expected to have a positive impact on the employee. Thus, his proposal is based on two aspects: the employee’s skills with the organizational structure; and the employee’s motivation.

In the same year, ^{McDuffie (1995)} argued that the inclusion of human resource practices could contribute to improving an organization’s economic performance through the incorporation of the following practices: work teams, problem-solving group, participation, rotation, decentralization, recruitment and hiring, compensation, status differentiation, learning and training. Meanwhile, for ^{Delery and Doty (1996)} HPWPs can be appropriated by the organization as a market-type system that focuses on recruiting personnel from outside the organization and as an internal system, oriented to an extensive socialization of activities among the company’s internal personnel. Some of these practices are: career opportunities, training, results orientation, profit sharing, job security, participation and job descriptions.

Similarly, ^{Pfeffer (1994};¹⁹⁹⁸) argues that human resources practices contribute to the development of a competitive advantage through the management of workers’ capabilities and commitment. This author’s proposal indicates seven practices which are: job stability, contracting, retribution, communication, self-directed teams, training and status reduction. According to ^{Camps and Luna-Arocas (2012)}, this HPWP classification enjoys great academic recognition and in addition, they have been strongly referenced in the literature (^{Marchington and Grugulis, 2000}; ^{Pascual, 2013}; ^{Pascual and Comeche, 2015}) due to their internal consistency and findings that provide general support in the effectiveness of the implementation of the 7 practices in organizations (^{Luna-Arocas and Camps, 2008}).

Among the most recent classifications is the one presented by ^{Para-González et al. (2019)}, who argue that proactive organizations link HPWPs with their business strategy and, in this way, mobilize the capacity of individuals towards the company’s objectives. Some of these practices are: selection, training, performance evaluation and compensation.

Similarly, ^{Miao et al. (2021)} argue that the components of HPWPs operate in synergy and therefore, the effect they will have within the organization will be characterized by efficiency and will be related to positive attitudes on the part of employees. Some of the practices used in this study were: extensive training, empowerment, results-oriented performance evaluation, competency management, employee selection, information sharing and reward management.

On the other hand, when it comes to measuring HPWP, the researcher is faced with two major decisions. On the one hand, what to measure, i.e., which practices are considered HPWPs, and, on the other hand, how to measure these practices. Regarding the first decision, the approach followed by most authors to measure HPWPs has been to derive the practices from previous studies (^{Macky and Boxall, 2007}). In other words, researchers have identified practices that have been commonly referenced in the literature and used them in their studies. For example, ^{Camps and Luna-Arocas (2012)} carried out an analysis based on selected research and chose to group practices following the indications of ^{Pfeffer (1998)}. Under this same theoretical orientation, ^{Sun and Mamman (2021)} made a selection of dominant practices in the Chinese and Western literature with which they were able to pre-validate an instrument to measure them in their research in SMEs.

It was also noticeable that some authors have resorted to the adaptation of existing classifications. For example, ^{Kloutsiniotis and Mihail (2017)} adapted ^{Delery and Doty’s (1996)} proposal composed of 7 practices to be developed in health care services in Greece and in employees of service and manufacturing companies in China, respectively. For their part, in the research recently developed by ^{Takeuchi, Way, and Wei (2018)}, the authors carried out the measurement of practices based on nine items from a literature review.

As can be seen, researchers have not used a standard measurement of HPWPs, which has led to the existence of a considerable number of instruments for their measurement. In view of this situation, this paper aims to adapt the questionnaire proposed by ^{Camps and Luna-Arocas (2012)} that contemplates the seven HPWPs proposed by ^{Pfeffer (1998)} and that, when applied, could contribute to what has been done to date. To this end, previous studies developed by the authors (^{Camps and Luna-Arocas 2008}; ²⁰⁰⁹) were analyzed and the applicability in other research was reviewed (^{Pascual, 2013}; ^{Pascual and Comeche, 2015}).

Now that the set of HPWPs to be measured has been agreed upon, it is time to address the second major decision, which corresponds to how to measure them. In accordance with the way in which HPWPs have been measured, most research has obtained information from managers. Several criticisms have been made of the need to include not only the response of HR managers, but also the perception of HR management. The need to include not only the response of HR managers, but also the workforce’s perception of the use of HPWPs in organizations, has been criticized. For ^{Riaz, Townsend and Woods (2020)}, employee perceptions are very important because they give meaning to the HR system and, therefore, they should be able to translate the message that managers want to convey through the implementation of HPWPs. To this, ^{Xi et al. (2019)} argue that several studies have documented that HPWPs that have been reported by employees tend to present a lower valuation than that reported by HR managers or directors. This could be a result of a limited working relationship, in addition to weak consistency and coherence in HR systems that has contributed to variation in the interpretation and application of HPWPs by managers and employees (^{Baluch, 2017}). In this same sense, for ^{Mierlo, Bondarouk, and Sanders (2018)} the perception gaps between the different actors that make up the organization may occur because managers have other objectives, which may lead them to understand the practices differently from what they were designed for.

Based on these arguments, it is important not only to measure HPWPs under a scale that promotes the comparison of results to generate a contribution to the area of study, but also to perform an analysis based on the response of individuals and managers in order to establish the differences that may exist in the perception of their implementation.

3.1. Sample and data collection

The scope of this research was descriptive and cross-sectional. The information was analyzed under a quantitative approach using descriptive statistics to establish the gap obtained between the HPWPs perceived by employees and the intention of managers when implementing them (intentional HPWPs).

For the selection of the sample, the business information registered in the Chamber of Commerce of Bucaramanga is taken as a basis, which for the year 2017 reported a total of 21,931 SMEs. Due to the difficulty of approa­ching this large number of companies, it was decided to use a convenience sampling taking those companies that had a representative number of employees (more than 10) and in which the research team had some contact to conduct the study.

Table 1 below presents the characteristics of the empirical study.

A total of 50 companies participated in the study, with a minimum of 10 employees surveyed in each of them. The 601 employees and the 50 HR managers of these companies were characterized. The profile of the workers who responded to the questionnaire was made up in greater proportion -45.26%- by people between 26 and 35 years of age. Slightly more than half -56.07%- were women, and 47.09% of those surveyed said they were single. A total of 44.42% said they had children and 3 out of 10 had college education. In addition, 38.60% of those interviewed stated that they had been with the organization for between 1 and 3 years. For their part, 42% of the HR managers who participated in this research were between 26 and 35 years old. Approximately 3 out of 4 of those interviewed were women. Sixty-two percent of those interviewed said they were married. Sixty-six percent said they had children, 52% had a specialist level of education and 30% had been working for the organization for between 1 and 3 years.

The data collection process was carried out in those companies that, after receiving the letter with the presentation of the purpose of the study, agreed to participate in it. The collection was carried out through two methods: face-to-face and virtual through a link.

3.2. Instrument for measuring high-performance practices

Following the indications of some authors (^{Khilji and Wang, 2006}; ^{Macky and Boxall 2007}; ^{Xi et al. 2019}; ^{Ali et al. 2019}; ^{Riaz et al. 2020}) who argue that, in the analysis of HPWP use, both the worker and the HR manager should be included, it was decided to collect information from two primary sources, taking the same instrument for both cases. The first source was the employees of the organization and in this case the HPWPs were called “Perceived HPWPs” and the second source was the HR managers whose results were called “Intentional HPWPs”.

The instrument used was adapted from the proposal made by ^{Camps and Luna-Arocas (2012)} with a total of 21 questions (example: One of the company’s values is stability; The company has objectively and clearly defined job profiles, etc.), using a 7-point Likert response scale, with 1 being “never” and 7 being “always”. Additionally, other demographic questions were asked, such as age, gender, marital status, among others, to characterize the sample.

4.1. Validity and reliability of the instrument

The reliability and validity of the questionnaire used in an investigation are quality indexes of a study. In this case, reliability was determined through Cronbach’s alpha, obtaining 0.958, which is considered acceptable within the established values. On the other hand, the validity of the instrument was determined, as ^{Camps and Luna-Arocas (2012)} did, through a confirmatory factor analysis (CFA) that allowed a check of the dimensions of the HPWP measurement scale using the AMOS statistical package version 22. Some of the most commonly used indexes to evaluate the fit of the model, according to ^{Escobedo, Hernández, Estebané, and Martínez (2016)} are: RMSEA (< 0.05), p-value (0.05), Chi-square (2 to 3 degrees of freedom and with limits up to 5) and the RMR (root mean square error) that measures the variances and covariances of the sample. Table 2 shows the fit indices obtained through the CFA that verify that the scale is valid, i.e., that the items that comprise it measure the construct called HPWP. In this case, the p-value is 0.00 and the RMSEA is 0.049, with values less than 0.05 being acceptable. The CFI is 0.774, considered acceptable considering that the chi-square over the degrees of freedom is located between values 2 and 3. In addition, the RMR is 0.219, a number close to 0, which can be considered a good fit. In the words of ^{Morata-Ramírez, Holgado-Tello, Barbero-García, and Méndez (2015)}, it can be used to measure the construct through such an instrument if the model presents an adequate fit through these indices.

Similarly, Figure 1 shows the factor loadings of the indicators that make up the HPWP concept and that show the correlation between them. For this case, they all exceed 0.7 which corresponds to the cross-correlation index obtained, ECVI 0.871, expressed by ^{Escobedo et al. (2016)}, who argues that the closer to one the higher the correlation.

Source: Authors’ own elaboration.

Figure 1 Items used to measure high-performance practices 

4.2. Descriptive results

As mentioned above, the sample of this study was comprised of 601 employees and 50 human resources managers from different economic sectors, which contributed to the fact that, through descriptive statistics techniques, the gap was established between the responses offered by the employees (Perceived HPW) and those of HR directors (Intentional HPWP). Next, the characterization of sample c is presented, where it can be seen that the majority of participating companies were from the commercial sector with 22% and the transport sector with 20%.

Table 4 shows that almost all practices have a mean close to 5, in the case of the employees’ perception, which in our scale corresponds to “frequently” and 6 in the case of the managers’ response, which is the Intentional HPWPs and therefore would be closer to “almost always”.

On the other hand, it is evident that the relative variabilities expressed from the coefficient of variation among employees are usually greater than among managers, which means that the opinion of managers is more homogeneous than that of employees. However, it is worth noting that the HPWP with the lowest mean according to employees is that of compensation, but that for managers it occupies the third position on the basis of the scores from lowest to highest. In other words, for an employee, compensation is not always in line with what it should be, but the manager considers that this is an issue that is going well within the organization.

With respect to the Intentional HPWPs, it can be seen that the lowest ranked is training, while for employees this practice is in a middle position, i.e., the manager thinks that training could be improved while the employee does not consider the training he receives to be one of the worst practices (it is 5th out of 7).

On the other hand, the HPWP with the highest average is communication, both in Perceived and Intentional, followed by the HPWP of status reduction, which occupies a second level for both groups. In short, for both employees and managers, actions focused on communication and reducing distinctions between members of a company contribute to generating value for themselves and for the organization.

Given the difference in means observed previously, it is worth asking whether this difference is statistically significant, for which an analysis of the difference in independent means has been carried out (see Table 5) based on the t-student distribution, which shows that the results of employees and managers differ significantly with a level of significance or probability of Type I error of less than 1% (p<0.01). This leads us to be able to affirm that the responses of both groups are different and, therefore, to continue with their analysis.