Comparative Analysis between Quality of Life and Human Labor in Countries Belonging to G7 and BRICS Blocks: Proposition of Discriminant Analysis Model

: The aim of the present research is to identify and analyze the variables which help to e ﬀ ectively di ﬀ erentiate Quality of Life (QoL) and human labor in the G7 (Germany, France, Italy, Canada, Japan, United Kingdom, and United States of America—USA) and BRICS countries (Brazil, Russia, India, China, and South Africa) through a discriminant analysis. A discriminant analysis model is developed to classify countries as having a low, mid, or high QoL based on QoL and human labor variables. The variables used in the discriminant analysis were obtained between 2010 and 2022 from two platforms: NUMBEO variables capable of relating QoL to socioeconomic aspects and OECD’s (Organization for Economic Cooperation and Development) human-labor-related variables. Based on the results, the three variables that most discriminate the groups in order of importance are employed women in relation to the female population, the female labor force participation rate, and the female unemployment rate. Countries are classi ﬁ ed as having a low, mid, or high QoL. The adopted technique will allow researchers and managers to classify and draw goals for action reorganization and investment in QoL and labor.


Introduction
Understanding the association between quality of life (QoL) and labor is essential to formulate public policies and business strategies that promote well-being and individuals' productivity (Klein et al. 2019).QoL and human labor are factors contributing to individuals' well-being and to countries' socioeconomic development.
The World Health Organization (WHO) defines QoL based on the world view each individual has over their place in life when it comes to habits, beliefs, behaviors, difficulties, and future uncertainties (WHOQOL Group 1995, p. 1405).Accordingly, if one has in mind the social context individuals are placed in, differentiated QoL measures must be taken based on the multidimensional concepts of well-being and QoL at work, which become essential at the time to establish a healthy physical and mental life (Blick et al. 2016;Boreham et al. 2016).
The concept of labor is understood as professional activity, paid or not, productive or creative, carried out for a given purpose (Neves et al. 2018).Work environments can encompass physical, remuneration, growth opportunities, and balance between professional and personal life conditions, among others.
Both concepts of QoL and labor are linked.QoL has an impact on labor conditions, and such a QoL perception by the population covers either individual subjective aspects or individual and socio-environmental factors that surround a given socio-cultural context, as well as aspects like objectively measuring human life conditions, healthm and labor (Bartolović 2024;Cecere et al. 2023;De Oliveira et al. 2021;Louzado et al. 2021;Randall et al. 2023;Razika et al. 2023;Sousa et al. 2023;Turgut et al. 2024).Different statistical methods and questionnaires have been used to investigate and analyze QoL and its relation to labor, such as WHOQOL-bref (Fleck et al. 2000), WHOQOL-100 (Fleck et al. 1999), cluster analysis (Bird et al. 2023), logistic regression (Maryam et al. 2023), and, most recently, discriminant analysis (Dunuwila et al. 2023).Discriminant analysis is a technique mainly applied as a classification method.It is often used to reduce resources' dimension and extraction (Deng et al. 2024;Imelda et al. 2024).Discriminant analysis allows researchers to analyze differences between groups by simultaneously taking into account different variables (Silitonga et al. 2021).Gori et al. (2023) adopted discriminant analysis in a case study about Italy to assess individuals' QoL based on the influences of perceived stress, anxiety, concern, and defense mechanisms, according to different post-traumatic stress levels.Khatatbeh et al. (2021) investigated managerial support levels that discriminate pediatric nurses in the following dimensions: burnout, QoL, intention to quit, and patients' adverse effects.Khatatbeh et al. (2021) compared stroke patients with different functional mobility levels to predict functional clinical measures capable of being combined in order to predict functional mobility.
Given the multidimensional natures of QoL and labor, using indices is a recurrent practice in research (Alaimo and Maggino 2020;Barrington-Leigh and Escande 2018;Peiró-Palomino and Picazo-Tadeo 2018;Simões et al. 2015).These indices are adopted as tools designed to measure society's features in order to ensure its sustainability and development (Bianchi and Biffignandi 2022).They are also used to analyze topics, such as human well-being, economic development, sustainability, real estate market, and life standards (Kuc-Czarnecka et al. 2020).Comparing social, economic, and environmental data among countries also depends on indices that help with measuring public policies, the so-called performance indices, which include unemployment comparisons between men and women (Becker et al. 2016).
Other variables are taken into consideration to calculate development indices, among them, Purchasing Power index, Safety index, Health Care index, Cost of Living index, Property Price to Income Ratio index, Traffic Commute Time index, Pollution index, and Climate index.These variables help to estimate the general quality of life of a country's inhabitants; they are disclosed by the NUMBEO platform, as shown herein (Shu et al. 2022).
Other studies, such as that by Sehnbruch et al. (2020), analyze the dimensional quality of jobs and QoL recorded for individuals in Latin American countries based on labor income indices, job stability, and conditions.Accordingly, the aforementioned authors created a quality of work index to identify the behavior of these variables in comparison to labor.Based on the global context, the G7 (Germany, France, Italy, Canada, Japan, United Kingdom, and the US) and BRICS (Brazil, Russia, India, China, and South Africa) are the two economic and political blocks playing significant roles in the world economy.Together, they account for 53% of the world population, 69% of the global GDP (gross domestic product), and 49% of the world trade (Castanheira et al. 2021).
The G7 countries are known for their high economic and social development levels and their established well-being systems (Abbasova et al. 2022).The BRICS countries house 3.2 billion people, they cover 29.3% of the total global surface, and they gather 45% of the global labor force, explaining their huge growth potential, since they represent 43% of the world population and 30% of the global GDP, whereas the G7 countries represent 10% of the world population and 39% of its GDP (Numbeo 2022b).
The aim of the present research is to assess QoL and human labor variables in a complementary way.This allows for discriminating factors related to labor influencing the QoL classifications in the G7 and BRICS countries.Such an analysis must consider the QoL indices recorded for the G7 members, which are higher than those recorded for the BRICS countries.Therefore, the aim of the present study is to identify and analyze the variables that help to effectively differentiate QoL and human labor in the G7 and BRICS countries through a discriminant analysis.The proposition is to develop a discriminant analysis model to classify countries based on a low, medium, or high QoL, according to QoL and human labor variables.

Methodology
The variables included in this study were selected from the NUMBEO and OECD platforms.The NUMBEO database provides information provided by 475,697 collaborators from 9161 indexed cities.It provides updated data about quality of life conditions.The platform has been registered in Serbia since 2009 under n.20853514 and its methodology is based on collecting information in a database system by, amongst others, user contributions provided by residents in the indexed cities and manual collection via surveys and other crowdsourcing and validation techniques in authorized sources (supermarket websites, taxi cab websites, governmental institutions, and newspaper reports) (Numbeo 2022b).
Similarly, the Organization for Economic Cooperation and Development (OECD) is an intergovernmental organization comprising 38 member countries, and one of its main goals is to provide information for political and governmental decisions aimed at seeking dialogue and cooperation.It is composed of a supervision and strategic guidance council encompassing ambassadors from the member countries and from the European Commission, as well as of discussion and review committees, experts and work groups for data analysis and assessing the political actions taken by OECD member countries, and of a secretariat of evidence and analysis.The 3300 secretariats comprise economists, lawyers, scientists, political analysists, sociologists, digital experts, statisticians, and communication workers.This platform's statistical database is split into topics, such as agriculture, economy, environment, finances, government, innovation and technology, job positions, and society, based on the 1956-2023 historical series.
Quantitative variables were selected in the NUMBEO and OECD online platforms.The selection criteria applied to the variables met their relevance and association with the QoL and human labor topics.These two platforms were selected for the study because they provide consolidated data on all countries and because they are recurrent researchdata sources (Carlsen and Bruggemann 2020;Demydyuk 2019;Guo et al. 2022;Lee and Chou 2018;Peiró-Palomino and Picazo-Tadeo 2018;Polzin et al. 2015;Saturno-Hernández et al. 2019;Shu et al. 2022).
The assessment of QoL and human labor is based on a discriminant analysis, supplemented by an F test, M test, and the Kernel Fisher discriminant analysis (KFD), to develop a validated prediction model for this study, which was employed as follows.
The variables used in the discriminant analysis related to the QoL and socioeconomic aspects selected from the NUMBEO platform were the quality of live index, purchasing power index, security index, health care index, cost of living, Property Price to Income Ratio index, Traffic Commute Time index, and pollution index.Additionally, the following human labor variables were selected from the OECD's platform: employed women in comparison to the female population, the rate of women participation in manpower, the rate of female unemployment, employed men in comparison to the male population, the rate of men participation in manpower, and the rate of male unemployment (for a summary, see Table 1).The present study encompasses data spanning the 2010-2022 period to construct a historical series and mitigate COVID-19's effects on the indices recorded between 2019 and 2022.Countries in both the G7 and BRICS were categorized into three QoL levels according to the data and classification intervals available on NUMBEO's platform (Table 2).As the information manually gathered from the aforementioned sources in our study is updated twice a year by the providers, we applied both automatic and semiautomatic filtering to remove data classified as 'noise'.Another filter helped to eliminate the lowest and highest quartiles of the data, as these extreme values are more likely to be erroneous.The lowest, highest, and medium values of the remaining data were calculated and reported.NUMBEO also archives old data values, based on a standard 12-month data discontinuation policy, however, it may utilize 18-month data when newer data are not available, or the indicators suggest low inflation in a particular country.Old data values are preserved to be used for historical purposes.
NUMBEO's QoL index does not reflect a causal value; rather, it represents the sum of all variables composing it, specifically: Purchasing Power index, Safety index, Health Care index, Cost of Living index, Property Price to Income Ratio index, Traffic Commute Time index, Pollution index, and Climate index (Numbeo 2022a).
Dependent variables regarding low, medium, and high QoLs were categorized into groups 1, 2, and 3 to allow for classifying countries based on similar discriminant features.The countries were categorized into a Low, Medium, and High QoL according to the historical average for the period from 2010 to 2022 for the QoL variable on the NUMBEO platform, as shown in Table 3.The identification and selection of indices as independent variables was based primarily on nomenclature, as shown in Table 4. Discriminant analysis was selected from various multivariate statistics as the method to establish relationships between the non-metric dependent variables and the metric independent variables.This approach identifies the variables most relevant to explaining the differences among groups that are divergent in a specific context.It optimally discriminates the variables in order to maximize intragroup variance.Accordingly, applying the discriminant analysis model helps to better understand the variables that effectively differentiate QoL and human labor in the G7 and BRICS countries.Two assumptions are crucial for the application of a discriminant analysis: the first assumption is based on multivariate normality in explanatory variables and on homogeneity in the variance and covariance recorded for groups.The second assumption concerns the homogeneity of the variance and covariance matrices across different groups, which can be assessed through Box's M statistics (Lourenço Niza and Broday 2022).Furthermore, F tests helped to determine if there were significant differences between the means of two or more groups by comparing the variances within groups to the variances between groups (cf.for example, Wang and Cui 2013).
Discriminant analysis helps to place variables within discriminant functions to analyze how they explain the models based on statistical weights resulting from the Wilks Lambda Test (Carmelino and Hanazato 2019).
The Statiscal Package for the Social Sciences (SPSS), version 23, was used to classify and analyze the data collected from the two above-described platforms (NUMBEO and OECD).This procedure comprised dependent (quantitative/nominal/non-metric) and independent variables (quantitative/metric) (Adeyemi et al. 2020).
Therefore, values 1 to 3 represented the groups corresponding to the quality of life scale according to NUMBEO platform's indicator.After classifying the groups, the Wilks Lambda test was used to test the hypothesis concerning the mean equality among groups or to determine if at least one of the groups differed significantly.This allowed for the testing of the following hypotheses: H0: mean equality among the low, medium, and high QoL groups.
H1: at least one of the low, medium, and high QoL groups is different.The Box's M test was conducted to assess the potential homogeneity in the variance and covariance matrices.The following hypotheses were assumed for this test: H0: matrices are homogeneous

H1: matrices are not homogeneous
The recorded self-values were analyzed to observe how different groups varied within discriminant functions; the greater the distance between the recorded self-values and 1, the larger the differences among the groups.These variations are explained by discriminant functions.The structure matrix was elaborated to show the correlation of each independent variable with each discriminant function.Variables marked with an asterisk (*) were the most relevant in determining each discriminant function.
The non-standardized coefficients of each independent variable enabled the formulation of discriminant functions for each of the three groups and for a description of the coordinates of the groups' centroids.Centroids are measurements applied to assess the typical points of individuals in a group and to graphically illustrate the distancing level between groups.The centroid coordinates for each group were obtained through the discriminant analysis, which highlighted the distancing level and individuals' typical points.Canonic discriminant graphics were plotted to help visualize the groups, showing individuals' distribution based on common features.
Last but not least, a KFD analysis was included, which is a non-linear classification technique that enhances Fisher's linear discriminant analysis.This approach maps input data into a higher-dimensional "feature space", where linear classification can be effectively performed, which corresponds to a powerful non-linear decision function in the original input space (Mika et al. 1999, p. 43).Similar to Georgiou et al. (2023), Kernel Fisher discriminant analysis models were employed to find two or more correctly separated classes or groups.This included different steps: (1) mapping to a higher-dimensional space, (2) computing the kernel-matrix, (3) between-class and within-class scatter matrices, (4) maximizing Fisher's criterion, (5) extracting discriminant functions, (6) classification function coefficients, and (7) classifying the new observations, which were supported through the SPSS data analysis.

Results
Our analysis shows that the variables (as described in Table 4) 'employed women in comparison to the female population', 'women participating in the labor market', 'rate of unemployed women', 'employed men in comparison to the male population', 'rate of men participating in manpower', 'rate of unemployed men', 'purchasing power index', and 'costs of living index' are highly correlated.However, the decision was made to retain them because excluding these variables would result in a loss of information in the discriminant function model.
Table 5 shows significant differences between the groups' averages in the following independent variables (p-values < 0.05): (X1) employed women in comparison to the female population, (X2) rate of women participating in manpower, (X3) rate of unemployed women, (X4) employed men in comparison to the male population, (X5) rate of men participating in manpower, (X6) rate of unemployed men, (X8) safety index, and (X11) property price in relation to income.In other words, the difference between the means recorded for the groups is significant when it comes to these variables.The lowest value recorded for the Wilks Lambda test points towards the most important variable for the discriminant analysis, in this case, the variable of women employed in relation to the female population.Although the other variables shown in Table 4 are not statistically significant based on the F test, it does not mean that they should be ruled out from the analysis to enhance the functions' discriminant power (because the F test is considered as robust) and present little sensitivity to non-adequate data in comparison to the prerequisites of the analysis of variance (De Luna and Guimarães 2021).The Box's M test was carried out to test the homogeneity of the covariance-correlation matrix (Table 6).When the samples' dimension is equal to (p > 0.05), the null hypothesis is accepted, and it points out no difference among samples' covariance matrices.In this case, the null hypothesis was rejected, meaning there were population differences between assessments.The sample size may have influenced the Box's M results, although non-homogeneity did not impair the other analyses.In this particular case, it was possible to proceed with the discriminant analysis (Brene et al. 2014).
The function of the canonical correlation is based on the concept of principal component analysis, which can simplify the differences between two sets of variables (Yang et al. 2023).Table 7 introduces self-value outputs and their canonical correlations.The variance rate in function 1 explains 77.30% of the data variation in the discriminant analysis.Self-value sets the superiority degree among functions: function 1 (selfvalue = 0.937) is superior to function 2 (self-value = 0.276).
Explanatory power is given by a squared canonical correlation, and it has the same explanatory power recorded for R2 in a regression analysis.Accordingly, the value of the squared canonical correlation in function 1 (0.696) 2 showed the mean explanatory power of the function (48.44%).Thus, the reliability degree recorded for function 1 was acceptable, but that recorded for function 2 (0.0465) 2 only reached 21.62%.Therefore, it did not reach the acceptable levels.
It is possible to conclude that 48.44% of the behavior recorded for the discriminant dependent variable was explained by the variables in the model, based on function 1.
The second Wilks Lambda test (Table 8) assessed the self-value's significance for each discriminant function: the lower the lambda value, the greater the difference among the groups' means (Reis et al. 2019).The results in Table 7 show that function 1 is highly discriminant based on Wilks Lambda (0.405 and p < 0.05).Table 9 introduces the structure matrix, showing the correlations of each independent variation with each discriminant function.Variables followed by (*) are the most relevant to determine each discriminant function.The variables of women employed in relation to the female population (X1), female labor force participation rate (X2), female unemployment rate (X3), safety index (X8), pollution index (X13), and care index with health (X9) have a higher correlation with discriminant function 1, while the independent variables of male participation rate in the labor force (X5), employed men in relation to the female population (X4), property price in relation to income (X11), daily displacement index (X12), purchasing power index (X7), and cost of living index (X10) have a higher correlation with discriminating function 02.The variable of male unemployment rate (X6) was not used in the analysis.
The centroid function of a group (Table 10) shows the cut-off point used to classify the cases.Cut-off points organize the discriminant values to relate events based on the groups.The dispersion graphic plotted for the groups' elements (Figure 1) comprises the values of the discriminant functions and the groups' centroid values.It is possible to observe the purple group (Group 1), the green group (Group 2), and the yellow group (Group 3).The groups are very close, which shows that the classified individuals have very similar characteristics.Some individuals in Group 1 are classified in Group 3, probably influenced by the variables of X7, X9, X10, X11, X12, X13 (Medium Quality Value), and 3 (High Quality Value), which are better classified, as this is not a fixed category, but rather a transitional one, in which countries move between one category and another or advance from one point to another.In a further analysis, we employed the KFD analysis model, which is a technique to find linear functions capable of two or more classes or groups (Georgiou et al. 2023).Table 11 introduces the KFD classification function coefficients used to classify the observations in each group.The KFD functions helped to classify a given case according to the highest value recorded for one of the three discriminant functions.Thus, the following discriminant functions exist for Group (G1), Group (G2), and Group (G3), with each variable represented by (X1) employed women in comparison to the female population, (X2) rate of women participation in manpower, (X3) rate of female unemployment, (X4) unemployed men in comparison to the male population, (X5) rate of men's participation in manpower, (X6) rate of unemployed men, (X7) purchasing power index, (X8) security index, (X9) healthcare index, (X10) cost of living index, (X11) Property Price to Income Ratio index, (X12) Therefore, it was possible to test the functions generated by the model proposed in this paper, which used QoL and human labor variables to predict the classification of the countries in the G7 and BRICS blocks based on a low, medium, and high QoL.The applied data were extracted from the NUMBEO and OECD platforms in 2022.Unfortunately, data for variables X1, X2, X4, and X5 related to China were not available or disclosed, which impaired the classification of this country.Table 12 synthesizes the data used for the variables in each G7 country.Similarly, Table 13 provides the data from 2022 which were applied to the BRICS countries.The highest values recorded for the functions' results (Table 12) indicated the group into which a specific country was classified (see Table 15).It is possible to identify that Japan and France were classified into Group 1 (Low QoL) after applying the model.This classification resulted from the fact that these countries have lower human labor indices compared to the other G7 members.No country within the BRICS bloc was classified as Group 3 (high QoL).

Discussion of the Main Results
The main objective of the present research was to identify and analyze the variables that can effectively differentiate QoL and human labor in the G7 and BRICS countries based on a discriminant analysis.The results showed that the variables most discriminating the groups, based on order of importance, were employed women in comparison to the female population, the rate of women participating in manpower, the rate of unemployed women, employed men in comparison to the male population, the rate of men participating in manpower, the rate of unemployed men, safety index, and property price in relation to income.The variable of women employed in relation to the female population recorded lower Wilks Lambda values.
The classification results were compared based on the prediction model.France and Japan were classified as having a low QoL, South Africa, India, Italy, and Brazil as having an average QoL, and Canada, Russia, Germany, the United Kingdom, and the USA as having a high QoL.The classification by the model differs from the classification of the NUMBEO platform, and it is highly likely that such a change happened due to the application of human labor indices to form the classification.The main changes in relation to the classification of groups between the model and the platform were: Russia, which went from a low QoL (NUMBEO) to a high QoL (model), the United Kingdom, which went from a medium QoL (NUMBEO) to a high QoL (model), and France and Japan, which went from an average QoL (NUMBEO) to a low QoL (model).
The other countries kept their classification into medium and high QoLs, in compliance with the index composed of QoL (NUMBEO).Thus, human labor indices in these countries did not reach an influence level high enough to increase or reduce their classification.
According to Shu et al. (2022), not all countries presenting a high GDP have a high QoL, and the Statistictimes (2023) ranking of global GDP sets the classification of countries based on GDP from the first to the tenth position, namely: US, China, Japan, Germany, United Kingdom, France, India, Italy, and Brazil.Although Brazil is classified as having a medium QoL, it emerges as one of the countries with the highest world GDP, because other variables influence this index's calculation, such as purchasing power index, pollution index, price/income ratio, cost of living index, safety index, health care index, traffic commute time index, and climate index (NUMBEO).A composite QoL index can summarize information in several variables, and it is much more accessible than trying to find out a common trend among them (Paruolo et al. 2013).
The aging of Japan's population is the highest in the world, and these high rates of aging combined with its low fertility rate have an effect on its GDP and play fundamental roles in both the reduction in the workforce and the saving population, although in economic areas, foreign capital investment mitigates this effect.The low fertility rate and the changes in the structure of the economy caused as a result of the aging of the population are directly related to indicators of deflationary pressure, an increase in unemployment rates, and a decrease in real GDP (Muto et al. 2016;Katagiri 2021).
In 2013, immigrant populations represented 8.9% of the French population, with more than 20% of immigrants being lower-level workers or administrative workers.Manual workers also tend to face job insecurity in France, with a third of them having temporary contracts, as permanent contracts are guaranteed for medium-and high-skilled jobs, and these are factors that contribute to the increase in unemployment rates in France (Bolet 2020).This may explain the change in the QoL classification according to the model presented in this study compared to the constant on the NUMBEO platform.
According to Olkiewicz (2022), Italian indices suffered from significant impact on GDP development because of the COVID-19 pandemic results-it recorded one of the most significant drops in this variable, which led to decreasing unemployment and socioeconomic indices.A recent analysis by Eurostat (2023) showed that the mean incomes of the Italian and French populations are the lowest in comparison to other countries in the G7, and these results are in compliance with those recorded in the current study, because both countries were the only ones in the G7 to present a medium QoL.
However, Bogoviz et al. (2021) assessed the social development contribution to economic growth in 2020 and highlighted that most indicators of the G7 countries, such as safety, health care, cost of living, traffic commute time, pollution, and climate indices, recorded a growth in comparison to previous years.BRICS countries, in turn, presented two indicators with negative links to the economic growth rate, such as cost of living and climate indices.This also helps to explains why all countries in BRICS were classified as having a medium QoL.
Other countries have highlighted the development of new indices models to measure economic growth based on the peculiarities of high-income and low-and middle-income countries, such as the G7 (Bethencourt and Kunze 2019;D'Aleo and Sergi 2017;Pereira et al. 2020) and BRICS countries (Pulido and Ustorgio Mora 2018;Sandberg et al. 2019).According to some studies (Bethencourt and Kunze 2019;D'Aleo and Sergi 2017;Pereira et al. 2020;Pulido and Ustorgio Mora 2018;Sandberg et al. 2019), development indices must also take into account the social and economic contradictions related to the need to improve employment and income indices.In addition, a recent study assessed the risks faced by older adults in China, Russia, and South Africa, and showed that these countries provide low food security to this population; this is why the trend of responses pointed towards their relative QoLs (Selvamani et al. 2023).
Our study focused on assessing the association between QoL and human labor in the G7 and BRICS countries, and will contribute to improving the knowledge about these blocks' socioeconomic and political conditions.It corroborates Al-Rabbaie et al. (2022), who analyzed governments' capital expense situations regarding production efficiency in the G7 countries.They found that these resources were invested in increasing physical production rather than improving socioeconomic outcomes, such as health care and education.Therefore, it is possible to observe that greater investments in production efficiency in the G7 countries led to higher employment and income indices, and, consequently, to an increase in the variables linked to QoL classification.In addition, the proposition for a discriminant analysis model led to a quantitative and systematic approach being applied to investigate these factors' influences on different contexts.
The prediction model introduced in this study allowed for a classification of all the countries on NUMBEO's platform.Similar research was carried out by Shu et al. (2022) based on linguistic quantifiers, aimed at comparing the NUMBEO's QoL index to the data analyzed in their research.Our model enabled slight adjustments to new QoL index scenarios so managers can recognize their own QoL level based on the conditions in their countries.
The structure model helps to present the association of each variable with each discriminant function (Tsehay and Shitie 2020).SPSS was used for practical applications.The set of variables in the applied model of joint independent variables showed a good rate of cases classified in the low, medium, and high QoL groups.The methodology employed in this study helped to identify the variables best discriminating the groups (Maroco 2003).Discriminant function 1 presented the best explanatory power degree: 77.30% of all data variation in the analysis.According to the Wilks Lambda test, the same function recorded a high discriminant power (p > 0.05 and Wilks lambda 0.405).Women employed in relation to the female population (X1), female labor force participation rate (X2), female unemployment rate (X3), safety index (X8), pollution index (X13), and care index with health (X9) were the variables presenting the closest correlation with function 1. Dickinson et al. (2023) used a discriminant function analysis for comparison purposes and to identify children with language development disorders (LDD).They recorded an 84.21% classification success based on the prediction model.Yadav et al. (2021) used the discriminant function to show that 83.1% of interviewees' results in India were properly classified into groups of individuals suffering from high blood pressure or individuals who did not suffer from it, as well as variations based on the socioeconomic context.Their results showed that high blood pressure prevalence reached 16.5% in the assessed population.
Another study that employed the same methodology was that by Thuany et al. (2021), which investigated the multivariate profiles of different types of Brazilian joggers who were successfully classified.The first discriminant function evidenced the variation groups for both the female and male sexes, with explanatory powers of 95.4% and 96.1%, respectively.This finding states the effectiveness of using the discriminant analysis methodology to classify elements and discriminate groups in studies with various objectives.
Finally, highlighting the limited range of studies presenting discriminant analyses related to QoL and human labor, as well as observing that the herein selected scientific studies introduce topics and aims different from those in the current study, emphasize that this is a relatively new area that warrants further investigation.

Policy and Practical Implications
Based on the results in this study, a few practical conclusions will be drawn in terms of public policy development, with the aim of assisting governments with strategies to increase the QoL indexes in their countries.
1.It has been shown that the variables most discriminating the groups, based on order of importance, were employed women in comparison to the female population, the rate of women participating in manpower, and others.
As an example, establishing policies that provide parental leave to alleviate the financial and mental burdens associated with motherhood and fatherhood promotes a balanced division of family responsibilities.For example, legislation implemented in countries like Sweden and Norway allows shared parental leave between fathers and mothers (Farré 2016).
Equitable access to education and professional training can also be made possible through scholarships and training programs.Investing in professional training can lead to significant improvements in women's employment prospects and wages, especially in sectors with low female representation such as engineering and technology (OECD 2022).
We also suggest promoting wage equality and career opportunities through regular salary audits, gender quotas in leadership positions, and mentoring programs.A study conducted by Ginglinger and Raskopf (2023) analyzed the effect of quotas for women on corporate boards and found that their implementation significantly increased women's participation in leadership positions, guiding companies towards social and environmental policies.
Implementing laws to protect female labor and combat gender discrimination in the workplace and aiming to prevent discriminatory situations such as harassment and discrimination in hiring and promotion with equal opportunity laws positively correlate with women in the workforce and equitable outcomes in the job market.Introducing microcredit programs to stimulate female entrepreneurship enables business expansion and income generation (Morazzoni and Sy 2022).
2. France and Japan were classified as having a low QoL, differing from the classification of the NUMBEO platform.Thus, the human labor indices in these countries did not reach an influence level high enough to increase their classification.
The aging of Japan's population is the highest in the world, and these high rates of aging combined with Japan's low fertility rate have an effect on its GDP, playing fundamental roles in both the reduction in the workforce and the saving population.
Given the higher survival rates of the current population, as well as improved health care conditions combined with a decreasing workforce, a series of reforms are being discussed and implemented in countries with low labor rates to retain the elderly population in the workforce.These reforms include flexible work schedules, enhancing the employability of older workers through the development of human capital, training and development programs coupled with technology, raising the retirement age, and improving working conditions, among other proposals made by the OECD (Nagarajan and Sixsmith 2023).Additionally, Husic et al. ( 2020) conducted a literature review aiming to examine publications on aging in the workplace and concluded that the most prominent articles suggested policies that encourage learning alongside a mixed workforce of young and elderly workers.
3. Although Brazil is classified as having a medium QoL, it emerges as one of the countries with the highest world GDP because other variables have influenced this index's calculation, such as purchasing power index, pollution index, price/income ratio, cost of living index, safety index, health care index, traffic commute time index, and climate index (Numbeo 2022b).
Although low-and middle-income countries are experiencing significant economic growth, many of them face challenges such as income inequality, persistent poverty, high mortality rates, unemployment, and lack of access to basic services.Implementing public policies to address these challenges can be crucial for ensuring sustainable and equitable economic growth, directly impacting the quality of life of the population (Milanovic 2019).
4. It is possible to observe that higher investments in production efficiency in the G7 countries led to higher employment and income indices, and, consequently, to an increase in variables linked to the QoL classification.
Last but not least, the practical and policy implications of this study allow for the identification of possible gaps that interfere with the Quality of Life (QoL) of the member countries of the BRICS and G7 blocs when work-related variables are inserted.The G7 member countries demonstrate satisfactory levels of public safety, income distribution, and labor, with a few exceptions regarding male and female participation in the workforce and employment rates, such as in the cases of France and Japan.

Conclusions and Study Limitations
Integrated policies that promote quality of life and labor are a great challenge for nations.Accordingly, monitoring and improving QoL, mainly its association with human labor, is becoming an issue widely discussed in several contexts.Many methods are used to measure public policies and development indicators to allow for decision making and the formulation of policies aimed at promoting social well-being.Thus, among all possible multivariate statistics, a choice was made herein to adopt the discriminant analysis, which is a technique used to set links between non-metric dependent variables and metric dependent variables, to help in identifying the most relevant variable to explain the differences among divergent groups within a given context that are homogeneous among themselves.
The aim of the present study was to investigate the variables best differentiating QoL and human labor in the G7 (Germany, France, Italy, Canada, Japan, United Kingdom, and USA) and BRICS (Brazil, Russia, China, and South Africa) countries through a discriminant analysis in order to propose a model capable of classifying countries with a low, medium, and high QoL in association with human labor.
Based on the results, the variables mostly discriminating the groups, in order of relevance, were employed women in comparison to the female population, the rate of women participating in manpower, the rate of unemployed women, employed men in comparison to the male population, the rate of men participating in manpower, the rate of unemployed men, safety index, and property price in relation to income.Based on the comparison to the QoL classification disclosed by NUMBEO's platform, only Canada presented changes in its classification due to the combined use of QoL and human labor indices.Its average jumped from a medium QoL to a high QoL based on the prediction model's classification.The study's limitations stem from the difficulties in locating indicators that were common across all countries and assessed over consistent time intervals.This limitation of the database may not accurately represent the realities in other countries for the purpose of comparing indicators.
Further research might employ other discriminant analysis techniques, such as a quadratic discriminant analysis and KFD analysis, alongside a larger data set and the application of countries' samples to reach a higher data heterogeneity in the selected variables.In addition, it is recommended to also include human labor qualitative variables in the research.The groups' homogeneity and the difficulty in locating Chinas' complementary data are also limitations of the present study.Finally, the significance of this study is evident in the development of a model capable of classifying Quality of Life (QoL) levels based on indicators of quality of life and human labor.

Table 1 .
Variables according to each platform.

Table 2 .
Interval of percentiles and groups' determination.

Table 3 .
Historical average for the period from 2010 to 2022 for the Quality of Life.

Table 4 .
Nomenclature of independent variables.
Source: Research data.

Table 5 .
Equality test applied to the groups' means.
Source: Research data.
Source: Research data.

Table 7 .
Self-values and canonical correlations.
a.The first 2 canonical discriminant functions were used in the analysis.Source: Resource data.
Source: Research data.
*. Highest absolute correlation between each variable and any discriminant function.b.This variable was not used in the analysis.Source: Research data.

Table 10 .
Functions based on groups' centroids.
Fisher's linear discriminant functions.Source: Research data.
Traffic Commute Time index, and (X13) pollution index:

Table 12 .
Classification data of countries in the G7 block.

Table 13 .
Data for classification of countries in BRICS block.
Source: Research data.

Table 14
presents the results in the prediction model after data were inserted in each QoL group's functions (low, medium, and high).

Table 14 .
Results of model functions.

Table 15 .
Classification based on the prediction model.