# Introduction pproximately 30 million workers are exposed daily to high levels of sound pressure in the United States only, with the potential for real damage to health. 1 Loss hearing induced by high levels of sound pressure is the second cause of occupational disease. 2 In 2006, only in Norway, of 3392 cases of occupational diseases reported to the government, about 59% were related to noise-induced hearing loss. 3 Checking occupational diseases most incident in Mongolia was observed the predominance of noiseinduced hearing loss, a phenomenon associated with the rapid industrialization that occurred in that country. 4 The magnitude of hearing loss would be a direct result of excessive exposure to noise, thus being dependent on several factors, among them those associated with exposure and related to the characteristics of the individual: sound pressure level (SPL), duration of exposure, type and frequency of the noise, susceptibility to damage by noise, age, history of hearing loss. 5 Models relating labor hearing loss to the behavioral of workers, named Health Promotion Model, analyze modifying factors (behavioral characteristics) and cognitive-perceptual factors (such as perceived benefits and self-efficacy). 6 The authors indicate that this and other developed models, allow them to infer that workers must be aware of the risk of noise-induced hearing loss (NIHL), and that they can do something to prevent of this disease. In a survey that took as sample a group of researchers in this area of occupational health and safety, noted that the notion of risk perception is socially constructed variable, being influenced by the nature of labor relations; valuesin society and individual events. 7 Although there are some publications on occupational noise exposure, the analysis of the individual perception of the worker and its implications on his behavior remains a poorly discussed subject. Moreover, the approaches verified relating the understanding of the behavioral phenomenon to the exposure to occupational noise, need more analysis and the support of quantitative approaches of main humans factors relacteds to NIHL. This paper aims to use the technique of factor analysis as a tool for examining the variable noise-induced hearing loss, to propose actions that may contribute to the prevention and anticipation of mitigating the risk of the exposed worker. # a) Theoretical Research For development of the theoretical research, there was conducted a literature search of indexed journals, seeking to identify main studies conducted in the identification of variables related to the development of noise-induced hearing loss. The research showed that the variables related to the development of NIHL would be influenced by a lot of factors related to the individual, as well as by the social and cultural organizational system, as shown in Table 1. Source: The author. For this research, a sustained quantitative methodology for factor analysis technique was developed, and a statistical model was adjusted to support the conceptual model.The proposed conceptual model, as shown in Figure1, is structured on the adaptation of similar models of other authors who also researched on this topic. 10,39,40 The self-administered questionnaires comprise the most frequently indicated version. 41 During filling, someone was available to assist in understanding the instrument, ensuring that no questions were left blank. X X X X X X X X X 9 X X X X X 10 X X X X X X X X X X 11 X X X X X X X 12 X X 13 X X X X X X X 14 X X X X X X X X X 15 X X X X X X X X X X X X X X 16 X X X X X X X X X X X 17 X X X X X X X X X X X X X X X 18 X X X X X X 19 X X X X X X X X X X X X X 20 X X X X X X 21 X X X X X 22 X X X X X 23 X X X X X X X 24 X X X X X X X X 25 X X X X X X X X X 26 X X X X X X X X X X X 27 X X X X X X 28 X X X X X X X X 29 X X X X X X X X X X 30 X X X X X X X X 31 X X X X X X X 32 X X X X X X X 33 X X X X X X X 34 X X X X X X X X X 35 X X X X 36 X X 37 X X X X X X 38 X X X X X X X X X X X X Total 828 In research on the influence of the perception of employees on the effects of occupational health hazards was also used as an instrument of data collection a questionnaire then generating similar analysis, with the objective of further reducing the dimensionality of the data. The analysis was useful for the planning of health promotion campaigns and prioritization of other interventions. 42 The initial questionnaire aimed to characterize the individual risk perception, consisting of questions divided into the following sections: worker´s identification (name, workplace, age, sex, marital status and education), individual risk perception and perception of noise effects, expectation and appreciation of the results of use of PPE, barriers, safety culture and risk behavior. The second questionnaire intended to characterize the noise exposure and the use of PPE, having questions grouped into the following sections: risk exposure, exposure to ototoxic chemicals, family history of hearing loss, training in occupational health and safety/hearing loss/use of PPE, comfort, and use of PPE and audiometry. The evaluation instrument used the Likert fivelevel scale, from "strongly agree" to "strongly disagree." After filling in all the 278 questionnaires, responses were converted into numbers, according to the scale, and tabulated in a spreadsheet. The answers were coded on scales from 1 (strongly disagree) to 5 (strongly agree). In some cases, the use of inverted scale was necessary, for example, for questions about barriers, physical load, and risk behavior. In a second stage, through the environmental noise assessment of each work micro-area of all evaluated employees, was possible the characterization of daily personal exposure level. The evaluated employees were still undergoing tonal audiometric examination by ISO 8253.1 standard. In a third stage, the model was tested and validated through multivariate statistical analysis which took into account, in addition to the construct (the development of NIHL), dependent subconstructs: risk perception; perception of effects; risk behavior and use of PPE. Because of this study include more than two dependents subconstructs, will be used the technique of multiple regression analysis (MRA), which, using dependent subconstructs whose values are known, allows predicting a single subconstruct independently Source: The author. Year 2019 G selected. 43 Equation 1 presents the MRA adopted in this article, which aimed to: 1) maximize the overall power prediction of a group of independent subconstructs as representatives of a composition; 2) to compare two or more groups of independent subconstructs regarding the predictive power of each factor. Y 1 =? 0 + ? 1 X 1 + ? 2 X 2 + ?+? m X m +? Equation 1 where Y1: Dependent subconstruct or criterion; Xi: Independent subconstructs or predictor; ? i : Regression coefficients; ?: Associated error; The path analysis is an essential resource of multivariate statistics, allowing correlations between characters are split into direct and indirect effects, measuring the influence of a variable, independent of the other, on the other. Path analysis was used to proof the conceptual model. Following the results of this analysis, the corresponding path diagram was established, as the summary presented in Figure 3. Statistical analysis was performed using IBM SPSS. # III. # Results Using the alpha (?) of Cronbach, it was possible to test the internal reliability, allowing a decrease up to 37,5% of the items. Thus, from the starting 74 questions, the technique applied in this questionnaire allowed to shorten it down to only 61 questions. 223 of the 278 employees evaluated in the sample were male and 55 were female, 147 had completed high school, 73 had incomplete graduation, and 44 had completed graduation. The estimated average hearing losses versus age, in the sample, may be observed in Figure 2, where the increase of hearing loss due to the aging of the population, by presbycusis, can be evidenced. The questionnaire also asked about the period of use and comfort of hearing protection equipment, aiming to quantify the subconstruct Use of PPE, in this case particularly the hearing protector was considereddue to the focus of the study on NIHL. As provided in table 5, of 278 evaluated employees, the vast majority -261 -reported using such equipment, and on average used their hearing protectors around 90% of the journey. The standard deviation was close to 18 points, featuring a small oscillation of the answers. From 261 people who reported using hearing protection, there was a divergence of habits between the sexes of 4% of usage time and 6% in the proportion. Chi-square test for equality has the value of 0,0976. Threshold for rejecting the hypothesis of equality at the 5% value is 3,84, one cannot drop the hypothesis of no difference between men and women in the PPE use. # a) Statistical Analysis # G Through the Pearson correlation coefficient matrix, shown in Table 3, it was possible to analyze the correlation between the subconstructs inspect. The correlation shows the linear relationship between two variables, where values will always be between -1 and +1. The sign indicates the direction and the size of the variable specify the strength of the correlation. An amount above 0,7 module is a strong correlation. Thus, variables with a rejected hypothesis of null relation at 5% level were considered significantly correlated. Source: The author. Table 3: Matrix of Pearson correlation coefficients between subconstructs Thus, considering the above, the research screen in this study uses the multivariate analysis of multiple regression analysis (MRA) data, using the data generated by the questionnaires. Five equations of multiple linear regressions were adjusted, with the After, the technique of path analysis was used, as the summary presented in Figure 3. This diagram symbolizes paths for subconstructs evaluated. Source: The author. IV. # Discussion # Table 4 shows the following results: a) Risk Behavior The study of the factor related to Risk Behavior of the sample has small effect on the development of hearing loss (0,013), where this parameter totally contemplates the indirect consequence arising from the interaction with the subconstruct Perception of Effects. This parameter may be positive, i.e., the higher the irresponsible behavior that puts the employee at risk of acquiring irreversible lesion in his auditory system, more significant these attenuations are. The Safety Culture, which contains only the direct component (0,099), presents a significant contribution to the construct parameter. As a positive value, it oddly reflects that the higher the presence of safe work procedures, the higher the level of hearing loss encountered. Perhaps this positive effect is justified by the conjuncture of the companies evaluated, which hada history of compromising workers' hearing and therefore implemented hearing conservation policies, such as HCP (Hearing Conservation Program). # c) Formation The study of formation did not contribute to the model since it presented no direct or indirect components in this evaluation. As for the characterization of this factor, we used questions to prior knowledge of the interviewee in areas such as safety and hygiene, as well as any previous training in the prevention of noise-induced hearing loss and about the correct use of earplugs and muffs. # d) Sex The factor that symbolizes the worker´s Sex has presented a significant value (-0,102), with a negative sign and contributed predominantly by the direct effect (-0,094) and the remainder coming from the indirect reaction through subconstruct Risk Index (-0,008). Its negative sign indicates that hearing loss tends to be lower in females and higher in the male sex, a fact which goes against table 7 (PPE use by sex). Regarding the sex distribution in the study sample, the vast majority of workers were male, a common factor in the industrial sector, and that may contribute to explain the results. I am giving a strong tendency for a higher incidence of hearing loss among females, when compared to the same sample of males. # e) Age The Age factor presents the highest of values (0,321), originated mostly by the direct effect (0,258) by the reaction of subconstructs Safety Culture, Risk Index, and Perception of Effects. It is of positive module, means that the higher the age of the employee, the greater is his degree of hearing loss. In this subconstruct study, the SPSS software used for the multiple linear regression automatically excluded the subconstruct Safety Culture, which seemed initially to collaborate indirectly to the composition of the factor. # f) Risk Index The Risk Index also contributes to the study of the components associated with occupational hearing loss, presenting a significant factor (0,136), with the direct and indirect effects (influenced by factors Age and Sex) having same values: 0,063. This parameter is related to the sound pressure level at work (dB), with the noise exposure time (years). The higher this indicator, the greater the average hearing loss of the worker. # g) Risk Perception h) Perception of Effects Regarding the Perception of Effects factor, there is a significant level (0,128) of participation, with a serious part being of direct effect (0,092) and the remainder (0,036) coming from the indirect interaction with the Age factor. As a positive number, it indicates that the higher the degree of perception of negative health effects for occupational noise, the higher is the degree of hearing attenuation. # i) Chemicals Ototoxic Drugs The impact study of Chemicals Ototoxic Drugs factor provides one of the lowest factors (0,014) of this model. Derived exclusively from indirect effects, it operates through the variables Age and Risk Index. As a positive value, it symbolizes the larger presence of these agents that are proven harmful to human hearing, the higher the occurrence of hearing loss of employees. # j) Use of PPE Finally, the impact study of the Use of PPE provides a discrete factor (-0,018) in this model. Coming exclusively from indirect effects, it operates through the variables Perception of Effects and Safety Culture. Negative value symbolizes the higher use of hearing protection devices -provided by the company to workers, the lower will be the occurrence of hearing loss in employees. V. # Conclusions Multivariate analysis was adequate and essential in the study of risk factors associated with occupational noise-induced hearing loss in the industry. There is a need for research on risk perception and mitigation strategies in the context of noise-induced hearing loss. Such is found tied to the increased probability of occurrence associated with the risk factors identified in the study. The development of risk management strategies to reduce NIHL requires both knowledge of the physical environment, and the social, psychological and economic processes that can affect The Risk Perception presents only one discrete factor (-0,030) of indirect effect arising from variables Age and Gender. As a negative number, it indicates the serious sensitivity of the employee against the risk of hearing loss, the lower will be the harm to the employee´s auditory system. The results collaborate to the trend that, the higher the perceived risk is (specifically through dimensions: Identification of risk sources, Knowledge about the noise, Perception of the efficiency of PPE and Means of protection), the smaller will be the chances of damaging the employee´s hearing. people's responses to environmental conditions of danger. Noise-induced hearing loss (NIHL) is a cumulative, insidious disease that grows over the years of exposure to noise associated with the workplacecausing damage to organ of Corti, usually bilateral, with progressive and irreversible loss and symptoms such as hearing loss, tinnitus, ear fullness, ear pain, dizziness, transient changes in blood pressure, stress, vision and mood disturbances, directly related to the exposure time, with sound pressure levels (SPL) and individual susceptibility. NIHL is caused by any exposure to a daily average of 85dB, for several years. Noise is responsible for about 20% of global hearing loss, including being the second leading cause of the occupational disease that affects more American workers. In an attempt to investigate the variables that impact on noise induced hearing loss, an analysis was developed taking into account the independent variable hearing loss, related to the variables dependent on metrics: Perceived Effects, Age, Risk Index, Culture of Safety and Sex. In order to predict any changes in the dependent variables -due to changes in the independent variable. The statistical analysis was applied to identify the latent dimensions of significant effect on the variable in the analysis of known risk factors, such as: Risk Behavior, Culture Security, Education, Sex, Age, Risk Index, Risk Perception, Perceived Effects, Chemical Substances, Ototoxic Drugs and Use of PPE. To test these associations were used Pearson correlation coefficient and multiple linear regression on the subconstructs identified. Through path analysis, was performed a factor analysis of direct and indirect factors related to NIHL effects. In this context, the present study examined the possible approaches that relate to understanding the exposure to noise risk in the workplace. The study enabled us to verify that Hearing Conservation Programs (PCA), and other campaigns, to be developed by companies, particularly in the industrial field, are based on modifiable factors identified. Thus, promoting actions that encourage the expansion of perception of the effects of hearing loss induced by occupational noise, will contribute to the prevention of the risk to workers' health. The path analysis showed how the exceptional impact variable on hearing loss, both direct and indirect effects aspect, the Age variable -demonstrating the effect of natural hearing loss related to agingpresbycusis. At the other extreme, the dependent minor indirect effect were variable Chemicals Ototoxic drugs. According to the results obtained in the present study, the individual perception of the risk of job exposure to noise is an important issue with regard to safe behavior, in particular, to avoid the involvement of occupational hearing loss by noise in the industry. Workers seem to avoid exposure to noise based on their perceptions of the risk. Unfortunately, it appears that they are bad appraisers this risk. Finally, other investigations may complement the study performed here. As this research was limited to the study of continuous or intermittent noise found in industrial environments surveyed, further research can evaluate the work environment whose employees are subject to impact noise. 1![Figure 1: Conceptual model proposed b) Applied ResearchIn a second stage, for the research development, data collection was performed using questionnaires, in a sample of 278 workers exposed to noise beyond the action level -defined in Brazilian law as 80 dB (A).The self-administered questionnaires comprise the most frequently indicated version.41 During filling, someone was available to assist in understanding the instrument, ensuring that no questions were left blank.In research on the influence of the perception of employees on the effects of occupational health hazards was also used as an instrument of data collection a questionnaire then generating similar analysis, with the objective of further reducing the dimensionality of the data. The analysis was useful for the planning of health promotion campaigns and prioritization of other interventions.42 The initial questionnaire aimed to characterize the individual risk perception, consisting of questions divided into the following sections: worker´s identification (name, workplace, age, sex, marital status and education), individual risk perception and perception of noise effects, expectation and appreciation of the results of use of PPE, barriers, safety culture and risk behavior.The second questionnaire intended to characterize the noise exposure and the use of PPE, having questions grouped into the following sections: risk exposure, exposure to ototoxic chemicals, family](image-2.png "Figure 1 :") ![source: The author.](image-3.png "") 2![Figure 2: Average hearing loss in the sample (age [years] versus loss [dB])](image-4.png "Figure 2 :") 1 Table2: Sample distribution between factoriesFactoryTotal number of workersNumber of workers in the sample (N)Number of workers in the sample (%)Age (years) Average Standard DeviationPeriod of service (years) Standard Average DeviationA68915222%36,210,115,78,6B5166312%42,99,23,34,8C1532013%48,211,18,211,6D2254319%42,68,97,99,5Total158327818%39,69,211,11,5 4VariableIndirect EffectDirect EffectTotal EffectRisk Behavior0,013-0,013Safety Culture-0,0990,099Training---Sex-0,008-0,094-0,102Age0,0630.2580,321Risk Index0,0610,0630,124Risk Perception-0,030--0,030Perception of Effects0,0360,0920,128Hearing Loss---Chemicals Ototoxic Drugs0,014-0,014Use of PPE-0,018--0,018 © 2019 Global Journals * Occupational noise and effects on blood pressure: exploring the relationship of hypertension and noise exposure in workers PPenney CEEarl AAOHN J 52 11 2004 * The Norwegian Labour Inspectorate's Registry for Work-Related Diseases: Data from David;Parker YogindraSamant Axel;Wannag EbbaWergeland International Journal of Occupational and Environmental Health 14 2006. 2014 Issue 4 * JamsranjavJavzmaa * ShagdarsurenOyunbileg Oyuntogos Lkhaasuren; * Analysis of Incidence Rates of Occupational Diseases in Mongolia Nyamjav;Sumberzul Wang Jung-Der International Journal of Occupational and Environmental Health 17 1 1986-2006. 2014 * A; DeCollee FBoodt EDegrave * Occupational exposure to noise and the prevalence of hearing loss in a Belgian military population: A cross-sectional study BGovaerts CLegrand PVVeken Noise Health 13 2011 * SallyLLusk * Comparison of the Original and Revised Structures of the Health Promotion Model in Predicting Construction Workers' Use of Hearing Protection OisaengHong DavidLRonis Research in Nursing & Health 29 2006 * Appraisal of Risk Perception in Occupational Health and Safety Research in Developing Countries TheaJHilhorst International Journalof Occupational and Environmental Health 2 4 2014 * Ricardo;Cordeiro Lima-Filho Euclydes C * Associação da Perda Auditiva Induzida pelo Ruído com o Tempo Acumulado de Trabalho entre Motoristas e Cobradores LilianC RNascimento Cad. Saúde Públ 10 2 1994 * References Références Referencias * História natural da perda auditiva ocupacional provocada por ruído SI CAlmeida De Rev. Assoc. Med. Bras 2000 * Percepção do risco de exposição ocupacional ao ruído PMArezes 2002 Tese submetida à Escola de Engenharia da Universidade do Minho para obtenção do grau de Doutor em Engenharia de Produção. Minho * Distribution of risk factors for hearing loss: Implications for evaluating risk of occupational noise-induced hearing loss MaryMPrince Journal of the Acoustical Society of America 112 2 2002 * Efeitos ototóxicos da exposição ao monóxido de carbono: uma revisão Adriana;Lacerda Tony;Leroux ThaisMorata Pró-Fono R. Atual. Cient 17 3 2005 * MárciaMedeiros Pinheiro De * Perfil audiométrico de trabalhadores do distrito industrial de Maracanaú -CE RenataTeles De Mesquita Rev. soc. bras. fonoaudiol 12 3 2007 * RaminMehrdad Mohammadi Saber; * Interaction of smoking and occupational noise exposure on hearing loss: a cross-sectional study GholamrezaPouryaghoub BMC Public Health 137 2007 * Interação entre grau de perda auditiva e o incômodo com zumbidos em trabalhadores com história de exposição ao ruído EDaniel HearingNoise ;Loss Ricardo;Cordeiro AdrianoDias A Review. Journal of School Health 74 6 16 2007. 2008 Rev. Bras. Otorrinolaringol. * KátiaAlvarenga JoséDe Freitas; Bastos Roberto Magalhães * Alterações auditivas em trabalhadores de indústrias madeireiras do interior de Rondônia. Rev. bras. Saúde ocup MagaliCaldana MarizaDe Lourdes; Feniman TatianeRibeiro; Jorge Martins; 2009 34 São Paulo * CarlaBotelho Silvana;Tomaz; Frota * Estudo comparativo de exames audiométricos de metalúrgicos expostos a ruído e ruído associado a produtos químicos AndréGonçalves Martins; Anna Paulla MaiaPaz Lopes * Revista Brasileira de Otorrinolaringologia 75 1 2009 * Percepção e o impacto da música na audição de integrantes de banda militar CláudiaGoncalves Giglio De Oliveira Al Rev. Soc. Bras. Fonoaudiol 14 4 2009 * CláudiaGoncalves Giglio De Oliveira * JairMarques Mendes * Ruído e idade: análise da influência na audição em indivíduos com Pedro Henrique DeMota Miranda * Pró-Fono R. Atual. Cient 21 1 2009 * MirAttarchi Saeed; MohammadMazhari Mahdi AmirMehrparvar Houshang * Cigarette smoking and occupational noise-induced hearing loss SaberMohammadi European Journal of Public Health 20 4 2009 * Abdullah;Ayçiçek FDereköy Sefa * Can Rh antigens be a risk factor in noise-induced hearing loss? FethullahKenar RamazanSargân EurArchOtorhinolaryngol 266 2009 * Análise de diferentes estudos epidemiológicos em audiologia realizados no AMArakawa MLCaldana Sales-Peres ÉIShc; Sitta Brasil. Rev. CEFAC 2010 * Efeitos auditivos em operadores de empilhadeira VanessaGarofani Adriana;Gregorczyk; Lacerda JairMarques Mendes; LeilaRibeiro Rev. Soc. Bras. Fonoaudiol 15 4 2010 * Gisele;Figueiredo Adriana;Lacerda * Achados audiológicos e queixas relacionadas à audição dos motoristas de ônibus urbano Massarolo Jeane;Neto JairMarques Mendes Rev. Soc. Bras. Fonoaudiol 15 2 2010 * A protecao auditiva utilizada pelos militares do Exercito brasileiro: há efetividade EduardoNeves Borba; MarciaSoalheiro Ciência e Saúde Coletiva 74 5 889 2010 * Patrícia Da SilvaChagas ThiagoCarlos; Diniz Hernandes * HeraldoGuida Lorena * Perfil audiológico em policiais militares do Estado de São Paulo SérgioKinoshita Koodi Arquivos Int. Otorrinolaringol 14 4 2010 * Noise induced hearing loss risk assessment in truck drivers. Noise and Health Ali;Karimi Kazerooni Farshid; Nasiri MohammadSaleh; Oliaei 2010 49 * Im Hj; Jang Tw; KimBg * The association between impaired fasting glucose and noise-induced hearing loss YJKwon J. Occup. Health 53 4 2011 * Hearing loss among classical-orchestra musicians. Noise and Health HeliKoskinen EskoToppila 13.50 2011 45 * Hearing impairment among workers exposed to excessive levels of noise in ginning industries. Noise and Health KamaleshDube Ingale SopanLalit; Ingale 2011 * Breaking News: Restaurants Serve Up an Extra Helping of Hearing Loss MatthewColeman The Hearing Journal 65 10 38 2012 * Determination of noise induced hearing loss in mining: an application of hierarchical log linear modelling Atakan;Mutlu MustafaOnder SeyhanOnder Environmental Monitoring and Assessment 184 2012 * Relative risk of elevated hearing threshold compared to ISO1999 normative populations for Royal Australian Air Force male personnel John;Attiaa May;Boggess MayaGuest Hearing Research 285 2 2012 * EmmaBrunskill * MarielFinucane * SethFlaxman * MayaMascarenhas * ColinDMathers * Global and regional hearing impairment prevalence: an analysis of 42 studies in 29 countries GretchenStevens EuropeanJournalofPublic Health Feb 2013 * CristianeOliveira Costa Da Cunha * Saúde auditiva e qualidade de vida em trabalhadores expostos a agrotóxicos Tereza Raquel Ribeiro De;Sena MarlizeteVargas Maldonado Ciência Saúde Coletiva 18 6 June 2013 Rio de * Hearing status among Norwegian train drivers and train conductors JohnsenEngdahl B Ts Lie A Skogstad M Tambs K Occup. Med 2013 * Luciana;Bramatti BetinaKunz Thomas * Relação da perda auditiva induzida por ruído e o uso de tabaco em trabalhadores de uma indústria alimentícia ElizianeMenin Gai Rev. CEFAC 16 2 Paulo Mar./Apr. 2014 * Health Promotion in Nursing Practice NJPender 1982 Appleton-Crofts Norwalk, CT * Training to promote worker's use of hearing protection: The influence of work climate factors on training effectiveness JBrady 1999 Michigan State University PhD Thesis * DartagnanGuedes Pinto * Reprodutibilidade e validade do Questionário Internacional de Atividade Física em adolescentes JoanaGuedes Elisabete CynthiaPinto; Lopes Correa RevBrasMed Esporte 11 2 Mar/Abr, 2005 * Mariano;Bonet Cañizares Mayilée; * Health-risk Perception in the Inner City Community of Centro Habana NiurysFernández Pedro;Mas RobertBTate AnnaleeYassi Cuba. International Journal of Occupational and Environmental Health 6 2014 Issue 1 * JFHair Jr * Multivariate Data Analysis RE;Anderson RLTatham WCBlack 1995 Prentice Hall Englewood Cliffs with readings. 4 ed