Validation of A Measuring Scale of The Fac- tors for The Employability of Millennials

Today the retention millennial workers is a challenge for organizations. The purpose of this paper is to propose ideas to the decision makers, to lessen the high rate of personnel turnover in all kinds of industries in many countries. This research validates an instrument that evaluates the employability factors of millennials in Mexico, according to their own expectations. We based our study on the previous literature about the millennial generation in many countries. To validate the questionnaire, a sampling of 781 workers from the states of Querétaro and Guanajuato in México was conducted. The method used to achieve the objective was through exploratory and confirmatory factor analyses. The confirmatory factor analysis was conducted using structural equation modeling, and tested two different methods: first order and second order models. With the three methods, exploratory, first order and second order factor analyses, similar results were obtained. In the analysis of the statistical techniques, two latent variables associated with the expectations of this generational group were generated. The factors found are “personal satisfaction with the organization” and “satisfaction with the organization’s social commitment.” These two factors are supported by the literature of other researchers. It is suggested that this questionnaire be validated in other countries but also in other regions of Mexico, using different productive sectors, thereby obtaining a broader perspective that will allow us to understand not only what millennials want from their work, but to what extent they want it.


Introduction
The purpose of this paper is to address the problem of the retention of millennials at work, by identifying the factors they want to achieve. Turnover costs for organizations decrease their productivity and competitivity. Voluntary turnover implies two different cost for organizations: separation costs and replacement costs (Allen et al., 2010). The separation costs are associated with HR managers expenses in the time devoted to paperwork, interviews, etc. Also, there are intangible aspects related to the lost of knowledge in the company. Replacement costs are associated with expenses for recruitment, selection and training. In many organizations there are special programs for attracting millennials as interns, but the results are questionable. Programs for employer branding and coaching are popular in a significant amount of global organizations in many countries like the US, Sweden, Brazil, India, México, etc. As we will present, researchers and practitioners are studying this phenomenon, but there are still some issues to be solved.
Millennials' retention is considered a challenging factor for human resources practitioners worldwide. Strategies such as "employer branding" or using multiple alternatives for online work have tried to meet millennials' expectations. On the other hand, researchers are evaluating different factors for retaining millennials at work (Stewart, Oliver et al., 2017;Thompson and Gregory, 2012). However, factors related to the commitment of millennials seem to have contradictory results (Thompson, 2011). In this paper, we establish factors for millennials' employability in México, from their point of view. We decided to adopt a particular focus for millennials, as Rudolph and Zacher (2017) suggested, taking into account that there are differences when defining a generation. We used the questionnaire developed by Ng et al., (2010) and evaluated it with a large number of students in Canada. In order to obtain our results, we used exploratory and confirmatory factor analyses. The exploratory factor analysis was conducted first to evaluate the number of grouping variables embedded in the questionnaire. The confirmatory factor analysis was conducted through structural equation modeling using Mplus software, to confirm the results obtained by the exploratory factor analysis. Finally, our results that there are two factors associated with what Mexican millennials expect from their work. These factors can be developed for the improvement of strategies by the human resources departments of the organizations and in this way contribute to the retention of millennials.
This paper is divided in five main parts. First, the paper states the literature review that contains the definitions of the terms used in the analysis. We segmented this part in a comprehensive way from generality to the specific, through reviewing the concept of the generation and its implications, the role that millennials perform in the labor context and its characteristics. Also, in this section the state of the art of the millennials literature and their expectatives at work are addressed. The second segment of the paper describes the methods used to meet the objective, the sample and the instrument utilized for gathering the data. In the third part, the results of the statistical methods utilized were discussed. The fourth part addresses the discussion of the findings of the results of the statistical outcomes. Finally, in the fifth part, the conclusion of the paper is presented.

Literature Review
This project aims to investigate what the millennial generation desires as remuneration and benefits when looking for a job. Human resources professionals have doubts when hiring people of this generation and what to offer to them, in terms of benefits and work conditions. Currently in organizations, there is a constant concern regarding the aspirations of workers classified as being from the millennial generation. These workers have different characteristics in relation to their aspirations within work and this has been reflected in the manpower turnover of organizations (Abou Assi et al., 2019). The above mentioned is not a minor problem and should be addressed to avoid the costs associated with this problematic situation. As an example, Tracey and Hinkin (2006) found, in the hospitality business, that there were significant costs associated with pre-departure, recruitment, selection, orientation and training, and finally the impact of a loss of productivity, but this also occurs in all types of organizations.
A multigenerational work place. Different generations normally coexist at workplaces in the labor market. According to Parry (2011) a generation is distinguished by two factors, the first by being at a common place in time and the second by having a collective memory. These two factors interact in a complex way to shape the life of any generation in different ways. This characteristic occurs during the early development years of the generation (Howe and Strauss, 2007). There are currently four types of generations coexisting at work. Table 1 shows the characteristics of these four generations. It can be seen that there are differences among the generations the workers belong to, in relation to their jobs and society.
Generational characteristics identify a part of the general population that was born during certain years and the significant events that occurred in those years influence the people's values and actions. For example, it is difficult to compare the work ethic of those born in the silent generation with that of the Generation X group (Schullery, 2013). Rudolp and Zacher (2017) suggested a lifespan perspective in order to explain the different generations. This approach assumes that cohort effects must be explained using a multidisciplinary method. In this sense, a person can be influenced by contextual factors which may influence the way that he or she behaves, independently of his/her age. Indeed, it is necessary to clarify that it is not only their age that identifies these groups, but rather their values and in this case their perspectives toward work.
Researchers are focusing their efforts on determining the differences between generations, especially for the factors related to work. Twenge (2010) found that: 1) Work's centrality and ethnicity is greater for the Silent generation , followed by the boomers and Gen X and lower for the millennials. 2) Altruistic values and intrinsic values, such as using talents, show no differences between all the generations. 3) Extrinsic values such as money and status are greater for Generation X, followed by millennials and in third place by boomers. 4) Job satisfaction and intentions to leave are greater for millennials and lower for Gen X. Twenge and Campbell (2008) found that the millennial generation had less need for social approval (dress codes for example), a higher self-esteem and narcissism than other generations, an external locus of control, higher levels of anxiety and depression, and an environment where women are more agentic and assertive. In another study Stewart and Bernhardt (2010) suggested that neuroticism and narcissism are increasing with successive generations, while self-confidence and achievement are declining. Of course, the differences described may impact in the workplace. Thus, the generational differences at work are a topic that is under development, considering also the need to incorporate new groups born after the millennials.

Millennials at work in the world and in Mexico.
According to the Financial Times (2019) there are 1,800 million millennials around the world. In China, millennials represent around 25% of the population. Nearly nine in every 10 millennials live in emerging economies. In general, millennials prefer to live in cities, marry later and postpone parenthood. Deloitte (2019) conducted a millennial survey in 42 countries and found that millennials showed a lack of faith in traditional societal institutions and are pessimistic about social progress, they are not particularly satisfied with their lives, they aspire to travel, help their communities and support companies that align with their values; many say they will not hesitate to end relationships when they disagree with companies' business practices, values, or political leanings. However, this description is broad and each individual or- Source: from Dries, Pepermans and DeKerpel (2008) ganization needs to know how to deal with millennial retention in order to promote productivity.
There is a consensus related to generational differences at work. For example, Mahmoud et al., (2020) found differences in the motivation to work between millennials and older generations in Canada. One particular case is that of the Chinese millennials, because they were born under the "only one child" policy implemented by the government at that time. This particularity caused millennials in China to receive unprecedented attention from their parents and grandparents. Zhao (2018) found, from a qualitative study, that companies should develop HR-related policies to better empower millennial employees; provide opportunities to develop their core competencies, create an altruistic working environment, and focus on millennial employees' job designs as well as career planning (p. 12). In an example conducted with Indonesian employees (Capnary et al., 2018), they found that the loyalty of millennials can be affected by providing flexibility in their work and life balance. In South Korea, a fast-growing economy in the past, the problems caused by millennials' complaints seemed to grow worse. Millennial employees were not willing to remain in a collectivistic culture as older workers did in Korean companies; they also refused to work long hours (Park and Park, 2018). These are some examples of millennial differences in the world, but there are more studies supporting ideas that millennials have unique characteristics (Costanza et al., 2012). In this case, we attempt to understand the millennial generation at work in Mexico.
In Mexico, millennials represent approximately a quarter of the Mexican population (almost 30 million), according to data from the National Institute of Statistics and Geography (INEGI). The statistical data of young people reported by INEGI (2011) are shown in Table 2. As presented in the table, there is an inequality in terms of work in relation to age and gender. Thus, we cannot state that the conditions of millennials are comparable across the world because not all countries have the same opportunities for education, employment and health. As we stated before, a person's behavior can be influenced by their social and economic situations and we believe the Mexican case is a good example of an emerging country.
In Mexican organizations, and also around the world, people work with other people who have different ages, ranging from the very young to those who are almost 80 years old. The age that people start work is mainly affected by socio-economic factors such as their family's income, while the retirement age has been prolonged because pension funds are facing shortfalls and many elderly people are unable to live exclusively on what they get from their pension (Mesa-Lago, 2019). This means that several generations of workers cohabitate together at work; these generations generally have common characteristics and are distinguished mainly by having lived through specific common events. These common events have changed the perspective of the lives of those born in certain years, for example: the Second World War, women's voting rights, sexual liberation, the Vietnam War, the creation of rock and roll, the fall of the twin towers in New York, the development of the internet, etc. This situation means that HR managers have to develop strategies to fulfill different needs, according to the age range of each person.
There is a high turnover in manufacturing plants, especially in the central part of Mexico. This phenomenon was investigated by Flores et al., (2019), but in the northern part of Mexico. They found that the turnover intentions of workers were related to factors such as burn-out and a lack of interest in the job. Industries that were prevalent in this studied area included car manufacturing, electrodocmestics , food, agrobusiness, etc. Human resources departments must implement strategies to retain their talented workers, but these strategies must consider the specific characteristics of the workers. The contribution of this project lies in specifying the characteristics of the millennial group of workers who will form the largest part of the labor force in Mexico, and around the world, in the near future.

Work engagement and millennials
As previously mentioned, the millennials' retention cost has an impact on most organizations, if turnover increases; organizational performance decreases (Park and Shaw, 2013). Van den Bergh and Wulf (2017) found that 61% of human resources professionals believe that millennials are hard to manage and incompatable with their workplace. Also, they found that six out of 10 employees belonging to this generation leave their place of employment within three years. Allen et al., (2010) argued that their salary is not the only factor in employees' turnover; key attitudes such as job satisfaction and organizational commitment are influencing factors as well. HR practices are essential for retaining employees, especially in the case of millennials. Providing potential new employees with an accurate description of their jobs will increase their motivation (Schroth, 2019).
Naim (2018) suggested that engaging with millennials represents a challenge. One of the strategies that he suggests is mentoring. Mentoring is a "process for the informal transmission of knowledge, social capital and psychosocial support perceived by the recipient as relevant to their work, career, or professional development" (Bozeman and Fenney, 2007, p. 731). Several studies indicate that the millennial generation pursues different goals at work. For example, Calk and Patrick (2017) conducted a study of university students and found that millennials are more collaborative than previous generations and will change jobs in search of free-time or a more challenging and satisfying work environment. Mahmound et al., (2020) suggested that, like older generations, millennials can be more effectively motivated through external rewards such as supervision, respect, and recognition from their subordinates.
Authors such as Walden et al., (2017), emphasized the relationship between the employee and the organization. These authors argue that when the organization has more information about an employee´s contribution to the organization, this gives that per-son a sense of commitment, and that organizational communication is positively related to job engagement and personal commitment. Recently, Lee and Li (2020) found that transparent internal communications have a strong relationship with the quality of the bond between the employee and the organization. Thus, the relationship between the organization and the millennial generation plays an important role and should be studied. Ng et al., (2010) conducted a study in Canada with a sample of 23,413 members of the millennium generation (born after 1980) with university students . These researchers found that this generation places great importance on the aspects of each individual's job, and they have real expectations about their first job, but they are looking for rapid progress and the development of new skills. An important aspect is that they seek a meaningful and satisfying life outside of work. In order to obtain these results, the following factors were taken into account: 1) Emphasis on the balance of life and work. Millennials give priority to "having a life" instead of "making a life." Events like September 11, 2011 have marked them. They have seen their parents work long hours only to become the victims of downsizing, layoffs and divorce. 2) Good salaries and benefits. The most important motivation of millennials is the emphasis on rewards and also on feedback. In this sense, there seems to be some dissociation between the abilities of millennials and their expectations regarding their rewards. However et al., (2019) found that, in Europe, young workers prefer a higher income and better job security due to the low quality of the working conditions. 3) Fast forward perspective.

Millennials expectations at work
Millennials seem to have high expectations when it comes to promotions and payments. They expect a promotion after six months at work (Erickson, 2009;Kowske et al., 2010). 4) Significant work experiences. Millennials expect much more in exchange for their work than just a salary. They look for jobs that are meaningful and satisfying (Yang and Guy, 2006;Hershatter and Epstein, 2010). 5) Enriched work environment. Millennials went to school where they were asked to work as a team. As a result, they emphasize the social aspect of work (nice co-workers, a fun environment), they wait for feedback from their supervisors. As previous mentioned, millennials' expectations at work differ around the world, and this issue could be addressed by investigating differences by groups, such as emerging countries, developed countries, etc.
However, the work done by Ng et al., (2010) needs to be proved by studying working millennials, because the shift from university to the work setting could change their perspectives. As Kuron, Lyons, Wchweitzer and Ng (2015) reported, there are differences that policy makers should evaluate. For instance, Williams (2019) reported that 98% of Irish employers were facing recruitment challenges and 41% of these employers were dissatisfied with the prospects available. In this sense, the emerging workforce is not meeting the managers´ expectations. HR managers nowadays are facing problems in recruiting and retaining a stable workforce. Arsenault (2004) mentioned that organizations will be less globally competitive if they fail to transfer the strengths of generational diversity to the workplace. Similarly, several authors agree that managers have a fundamental role in trying to detect how and why differences between the generations affect competencies, behaviors, attitudes and other attributes, as well as in implementing the best strategies to achieve the goals and objectives, and achieve a generational fusion (North and Fiske, 2015;Urick et al., 2016). The intergenerational conflict should be addressed using a holistic point of view through achievement-oriented strategies which include different perspectives in work settings, as well as differences in technology, people's abilities, communication skills and values (Urik et al., 2016). Therefore, we need to understand the requirements expected by millennials at work to create a better organizational climate in organizations. But, HR managers also need to fulfill the expectations of older generations as well.
We are in the era of "talent searches," companies have various programs to attract young people and develop their skills within the organization. Berthon et al., (2005) developed an employee attraction model taking into consideration five factors: interest value, social value, economic value, development value and application value. However et al., (2017), in their study of millennials in the United Kingdom, suggest suggested that the expectations of millennials should be reviewed and stated during the pre-employment stage to make them realistic and in line with the organization's HR practices. In their work they also suggest that national and contextual factors should be considered. Consequently, there is still space to contribute to the theory of the millennial generation's expectations from work.

Methods, Measurements and Sampling
To address the research problem and contribute to the existing literature, this paper proposes the validation of a measurement model that comprehends the key elements of what millennials want from their jobs. As stated in the literature review, there are still questions to be solved that professionals and academics need to allow them to fully understand these new kinds of workers. This project is intended to contribute to the orga-nizations´ productivity by reducing the costs associated with employee turnover.
The methodology used for the project was quantitative and included three different multivariate statistical techniques to validate the questionnaire. The main objective of this project was to determine what millennial workers want from their jobs through a proposed and validated questionnaire. The evaluation instrument used was the Ng et al., (2010). However, this scale hasn't been evaluated in the literature by employees, it was validated by students. The questionnaire was translated into Spanish and validated by human resources experts from two manufacturing companies. An exploratory study was not conducted because the human resources managers agreed with the items in the questionnaire. In the review with experts, one item (number 13) was removed because they considered it to be redundant. The sample was taken for convenience because only 20 companies participated and contributed information. However, a big sample was assured, which would meet all the requirements for the statistical validation techniques.
The research objective of the project was to validate the scale of the millennials' employability factors, for which a sample of 781 workers' data was collected; they ranged in age from 18 to 37 years old, and were from the states of Guanajuato and Querétaro (located in the central part of Mexico). The sample was collected through 20 different organizations that were interested in the turnover problem. These organizations were contacted by various universities and the authors of this paper. The sample was obtained by convenience with the organizations that participated in the study. The sectors that participated in the study were the manufacturing sector (seven organizations), health sector (three organizations), education sector (two organizations), the service sector that comprehend banks, insurance, etc. (three organizations), commerce sector (three organizations), government sector (one organization) and the agriculture sector (one organization). This sample is a good example of the number of organizations settled in the Bajio region of Mexico. In that region, the main businesses are the manufacturing plants for cars, owned by such companies as Mazda, Honda, Ford and VW, and their parts suppliers. Also, due to the climate, the region produces and exports vegetables and berries. Therefore, it is a cross-sectional, exploratory and explanatory study.
The sample comprised of the following: 58% were men, 42% women; in reference to their ages, 21% of the sample was aged from 18 to 22 years old, 19% was 23 to 27 years old, 46% was 28 to 32 years old and finally 14% of the sample was 33 to 37 years old. With respect to their schooling: 44% of the sample had a bachelor's degree, 39% finished high school, 13% held a master's degree, 3% had technical/professional qualifications and 1% doctorate. In relation to their work experience, the sample's members had the following characteristics: 62% had less than 5 years' experience, 27% had 6 to 10 years' experience, 9% had 11 to 16 years' experience and 2% had between 17 to 22 years' experience. The manufacturing sector employed 45% of the sample, 12% were in services, 12% in commerce, 13% in the health sector, 8% in the education sector, 7% in government and 3% in the agricultural sector.

Results
In order to validate the questionnaire and address the objective of this research, we followed four steps, we: 1) Calculated Cronbach's alpha and the Pearson's correlations between the variables. 2) Conducted an exploratory factor analysis with SPSS software.
3) Performed a first order confirmatory factor analysis with the Mplus software and calculated the Dillon-Goldstein's rho (or the composite reliability) which can be used for checking the internal consistency of indicators for each latent variable. 4) Calculated a second order confirmatory factor analysis to validate the dependent variable, the millennials' employability.
As a preliminary step, we tested the validity of the questionnaire. The Cronbach's alpha of the 15 items was first calculated and a value of 0.919 was obtained. Subsequently, Pearson's correlation of the 15 items was calculated, as shown in Table 3. The table also shows the mean and standard deviations of all the items considered in the questionnaire. As can be seen, all the items have a high positive correlation, indicating that there is a relationship between the questions.
As a second step to validate the questionnaire, we performed an exploratory factor analysis. The objective of this multivariate method was to reduce the number of variables into a common factor. To determine the adequacy of the exploratory factor analysis test, the Kaiser, Meyer, Olkin (KMO) sample adequacy measure was calculated, which contrasts whether the partial correlations between the variables are very small. It allowed us to compare the magnitude of the correlation coefficients observed with the magnitude of the partial correlation coefficients. The KMO statistic should be between zero and one. Small values indicate that the factor analysis is not adequate. In this case, as shown in Table 4, the values obtained in the KMO test are close to one.  188 Additionally, Barlett's sphericity test was calculated (Table 4). The test contrasts the null hypothesis, which states that the correlation matrix is an identity matrix, in which case there would be no significant correlations between the variables, and the factorial model would not be relevant. In this case, the significance value of the test was 0.00 which meant that the variables were highly correlated and provided a basis for the factor analysis (Leech et al., 2013). As a result of the exploratory factor analysis we concluded that the items could be grouped in two factors.
To complement the exploratory factor analysis, the eigenvalues, the load of the items, the percentage of the variance explained and the commonality of the items were calculated. This study was based on the varimax rotation analysis to evaluate the structure of the 15 items of the questionnaire. It is noted in Table 5 that the eigenvalues of each factor must be greater than one, since that is the generally accepted criterion. If the eigenvalue is less than one, it implies that the factor should not be considered. The variance explained by the first factor was 27.5% and that of the second factor was 26.3%, which totals 53.8%; in other words, the questionnaire explains 53.8% of the variance and the remaining 46.2% is explained by other factors. The communalities represent the relationship between the  variables and the rest of them before the rotation. If they are less than 0.30, they should not be accepted (Leech et al., 2013). Finally, all the factor loads were greater than 0.40, which measures how they contributed to the group or factor of which they are part. Table 5 shows that the items are grouped into two factors. The first one has been called "Personal satisfaction with the organization" and the second, "Satisfaction with the organization's social commitment." Items 1 to 7 are grouped into the first factor; the remaining items are grouped into the second factor. The third phase, or the analysis, tested the validity of the model. A first-order confirmatory factor analysis (CFA) was performed with the structural equation model using Mplus Version 7 software. The CFA allowed us to compare if the data were consistent with the theory (Verdugo et al. 2016).
To avoid any lack of normal data, tests were performed using the robust maximum likelihood (RML) method, as proposed by Satorra and Bentler (1994). The RML method represents an improvement on the exploratory factor analysis performed with SPSS, as it provides more accurate results. Table 6 shows the values of the standardized factor loads and the explained variance (R2) of the first-order CFA. It was observed that, for all the items, the exposed values met the criteria established for this technique.
Once the first-order CFA was carried out, the goodness of fit was assessed by determining the corresponding statistics, as shown in Table 7. The chi-squared statistic (χ 2 ), the root of the average quadratic residue, the root mean square error of approximation (RMSEA), the comparative goodness of fit index (CFI), the Tucker-Lewis index (TLI) and the standardized residual mean square root (SRMR) were calculated. All these goodness of fit parameters are generally used when presenting a SEM model (Vargas, 2019), so it was observed that the acceptance criteria were met, so the model had the required quality. The Dillon-Goldstein's rho for the first factor was 0.86 and for the second factor 0.83. These parameters are consistent with the Cronbach's alpha, which had been calculated previously. The results are shown in Table 7.
Subsequently, a second-order CFA was carried out in order to establish the relationship between the factors and an indicator, which in this case was the employability of millennials. A second factor CFA was conducted to corroborate the findings of the study. Standardized factor loads and goodness of fit statistics are shown in Figure 1.
The parameters of the second-order CFA showed a slight improvement over those obtained in the first-order CFA. Thus, the understanding of the theory presented in the measuring instrument was satisfactory for the statistical analysis. In addition, the goodness of fit measurements showed that the quality of the model was adequate.

Discussion
The research strategy chosen to validate the questionnaire used different statistical techniques developed for different purposes. The purpose of the analysis of the exploratory factor, according to Leach et al., (2013), is to represent a large number of the relationships among normally distributed variables, or the scale variables in a simpler structure. It is the first step in determining how to cluster the items in a construct. This step was conducted using SPSS and satisfactory results were obtained. The second part of the statistical study was conducted by confirmatory factor analysis, in which the specific model was tested and the relationship between the observed variables (items) and the conceptual constructs were evaluated. At first, we used AMOS to probe this model, but this software did not have the maximum robust likelihood method that corrected the bias due to normality. Thus, we decided to use Mplus developed by Muthen and Muthen which is more accurate for testing the parameters, but is also challenging due to the programming requirements (Muthen and Muthen, 2017). With the Mplus software we performed a first and a second order confirmatory factor analysis to test the differences in the structural equation models. We concluded that there were no significant differences but the second order factor model had better goodness of fit indicators.
In our tested model, Factor 1 was related to the "Personal satisfaction with the organization" while the remaining items were associated with the second factor "Satisfaction with the organization's social commitment." We confirmed that through the use of statistical techniques, it was demonstrated that the theoretical constructs proposed can be used to formulate strategies for organizations. This paper contributes by identifying the factors to retain Mexican millennials in their jobs. Human resources managers should regard these key issues when developing strategies in their organizations.
In general, these results confirm the values associated with the millennials proposed by Berthon et al., (2005) and Kowske et al., (2010). In our study, the sample was diverse, using early and late millennials, and we found there is still time to adjust the organizational policies to avoid the turnover problem. Our results demonstrate that Mexican millennials desire a personal stability and good working conditions, but they are also committed to the impact that they and the organization they work for have in the community. Working conditions in Mexico are not the same as in the US or other developed countries. Wages are a pending task for organizations and government policy makers (Samaniego, 2018). In this sense, we contribute to the literature by filling the gap in the theory of the millennial generation at work in emerging markets.

Conclusions
The main objective of this project was to determine what millennial workers want from their jobs through their replies to a valid questionnaire. This objective was reached through a rigorous statistical process. The contribution of the literature was in the development of the questionnaire about millennials' expectations, which was originally created by Ng et al., (2010); however we proposed two grouping variables: personal satisfaction with the organization and satisfaction with the organization's social commitment. Thus, this study confirms that millennials have a holistic perspective in relation to their requirements in the workplace. They do not only pursue material goods; they seek a full life within their personal and work environment. Our paper contributes to explaining, from the millennials' point of view, their requirements to remain in a job.
Authors such as Dries et al., (2008) indicated that there are four generations coexisting in the work environment and that each generation has different characteristics. As we can see from the results, the millennial workers seek to have a personal impact, a work-life balance, to engage with vulnerable groups and also with the diversity of the other employees. In this sense, it could be assumed that all the educational efforts to respect and improve our environment have had results because these programs were not taught to previous generations. Topics such as global warming, pollution, respect for the environment and human rights are present in the millennial generation's minds. The challenge for HR managers is how to use this global movement to generate strategies that can retain these employees.
The results of our project are also consistent with the results of Stewart et al., (2017) and Zabel, et al. (2017) as they all concluded that millennial workers do not differ significantly from other generations of workers; rather their approach is inclusive since they take into account some other aspects of the work. For example, from their point of view, an important factor is the performance results, but they do not measure productivity in terms of the hours worked but rather by what is achieved. The form of communication is different since they seek more direct communication, as well as frequent feedback from their supervisors (Gursoy et al., 2008). As the millennial generation will soon form the predominant part of the labor force in the world, organizations should take in account that the communication process has changed and that there should be a better understanding between generations, in order to gain organizational identification and satisfy the requirements of all the generations that actually contribute to the labor market.
As a recommendation to Mexican employers, based on our results, it is necessary that policy makers consider the millennials' expectations in order to decrease the high rates of rotation that currently exist in organizations. As described above, the workforce that is currently being integrated into organizations has different traits and that leads to changes in organizations, especially for their human resources departments. For example, Zavala-Villalón and Frías-Castro (2018) conducted a qualitative study in Chile by interviewing human resources managers and concluded that, in their opinion, millennials are autonomous, proactive and passionate about change and labor flexibility. Cuesta (2014) conducted a survey related to work in eight Latin American countries and found that, for millennials, the organizational climate, compensation and a healthy relationship between family and work are important retention factors. In this sense, we found that Mexican millennials should be engaged with the organization for them to remain at work. This implies that the HR managers should reflect about what are they doing in relation to creating environments where millennial employees feel satisfied.
The generation being studied presents a challenge for human resources strategies without a doubt. According to Naim and Lenka (2018), leadership strategies should be developed to help millennials act as mentors, and also include social networks in the workplace. Naim and Lenka (2016) also mentioned that, in order to develop an organizational commitment and remain in employ-ment, organizations must create professional development strategies. In addition to this, human resources managers should not lose sight of the fact that they must work with the varied generations that are currently in the organizational sphere (Rauvola et al., 2018). There is an urgent need to redesign the working conditions of employees to achieve their expectations (Williams, 2019), although these conditions should be restructured to create a better-quality working life, especially in emerging and underdeveloped countries.
This paper contributes to fill the gap in the literature of millennials' retention in emerging markets by proposing elements for their retention at work. The results of our study show that the level of engagement between the employees and their organizations contributes to fulfill the expectative of the millennials. The relationship between the supervisor, the group and the employee should be developed to promote a climate of trust and progress. The organization must provide the millennial worker with opportunities to develop new skills and abilities but also, he/ she must be satisfied with his or her salary. Also, the organizations should be interested in contributing to vulnerable groups and the environment. It could be interpreted that the implementation of the results of this paper could increase the costs for an organization, but the CEO must think that these costs are not comparable with the costs caused by the employees' turnover or layoffs. In many circumstances the lack of the required talent decreases productivity and increases costs in the good or service offered to the market. Human resources managers need to analyze their working environments and conduct actions to retain and develop millennial talent to contribute to increasing the productivity of the businesses.
Nowadays, with the COVID-19 pandemic crisis, the International Labor Organization (ILO) estimates that the loss of employment will be significative all over the word. This international organization estimates that during 2020 around 400 million full-time jobs will be lost. The most vulnerable in this situation will be young people and women. The most affected sectors will be hospitality business, transportation and manufacturing (ILO 2020). In this sense, HR managers have to work hard to retain good workers. In a significant amount of emerging countries, the crisis will be worse than expected, so social entrepreneurship could be an alternative for reducing this economic catastrophe

Limitations of The Study
Our research has limitations related to the sample. Although we used a large sample of millennials, they all came from the same region in Mexico. There are some other highly industrialized areas like the one we analyzed, in the northern part of Mexico. As future lines of research are derived from this study, it is necessary to inquire about the relationship between millennials' expectations and retention factors. Organizations must take care of these aspects to avoid hiring and firing costs and also retain talent within the organizations since this is the most precious asset that is available to create competitive advantages.
The challenge to reduce the high rates of turnover in many economies around the world, remains unsolved. This problematic situation impacts directly on the productivity and competitivity of organizations. As per the analysis conducted by Hom et al., (2017), the factors about why an employee remains at work have evolved to consider different factors and perspectives from those given by previous generations. Our research could be a starting point to consider turnover intentions being related with some other factors, as part of the employee-organizational relationship and job embeddedness (Mitchell, et at, 2001). Also, most of the studies related to turnover intentions are measured at a specific momentum of time. Academics need to explore them from a longitudinal point of view to see if there is a relation between the intention to remain and turnover, over time, and what factors are implied in the decision process of the employee. Other researchers could extend our research to other emerging markets and also relate our results with other variables, such as the working conditions or satisfaction with the supervisor. For Mexican researchers we suggest performing the study in other parts of the country, and in different sectors, to identify any differences and if there are, to what extent they have an effect, and in this way contribute to promoting the productivity in different industries.