Women in nursing teams : organizational identification and experiences of pleasure and suffering 1

1 Supported by Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), process # 402304/2010-1. 2 PhD, Adjunct Profesor, Escola de Enfermagem, Universidade Federal de Minas Gerais, Belo Horizonte, MG, Brazil. 3 PhD, Full Professor, Escola de Enfermagem, Universidade Federal de Minas Gerais, Belo Horizonte, MG, Brazil. 4 PhD, Adjunct Professor, Universidade Fundação Mineira de Educação e Cultura, Belo Horizonte, MG, Brazil. 5 PhD, Full Professor, Faculdade Novos Horizontes, Belo Horizonte, MG, Brasil. Objective: to analyze the degree of influence of the organizational identification on the experiences


Introduction
Identity is built through the performance of roles, in a given social context, and mediated by the identification process, in which the 'other' plays a relevant role.
Identity refers to the subject's self-perception in the interaction between the 'I' and society and between the interior and the exterior.Identification, in turn, relates to the subject's achievement of the qualities of the 'other', with a view to raising his self-esteem.
Through a set of identifications, the individual not only acknowledges 'who he is', but also gains the capacity to acknowledge himself as distinct, to perceive similarities and to understand collective life (1) .Identification takes place when the subjects' beliefs about the organization turn into a reference for his self-perception, generating a sense of belonging (2) .
The subjective process of the individual's bonding with the organization and with work can result in experiences of pleasure and suffering.For work to be perceived mainly as a source of pleasure, the individual needs to receive proper recognition while accomplishing tasks (3) , deriving from two sources of judgment: utility and beauty.The first relates to the technical, social or economic relevance of the activity the worker performs, as judged by a hierarchical superior; the second refers to the singularity and specialty level of the work performed, aiming for recognition by evaluating colleagues.Identification with work and the organization will derive from the results of these judgments (4)(5) .
In that context, the general aim in this paper is to analyze the extent to which self-perception and heteroperception influence organizational identification and the experiences of pleasure and suffering of female nursing team members at two general hospitals in Belo Horizonte/MG.Historically, nursing has been characterized as a predominantly female and numerically representative profession in the hospital context (6) .Care actions focused on people in nursing teams' diagnostic and therapeutic processes mobilize a great emotional loading and can generate experiences of suffering at work.Hence, the relevance of this study relates to the analysis of whether

Sample selection
Based on the list of female nursing professionals at the two hospitals ("A" and "B"), the functional distribution of the employees was obtained, with 229 baccalaureate nurses and 1038 nursing technicians, as follows: 142 nurses and 614 nursing technicians at hospital "A"; and 87 nurses and 424 nursing technicians at hospital "B".With a 5% error margin and a 95% confidence level for finite population, the total sample was estimated at 511 subjects, including: 61 nurses and 224 nursing technicians at hospital "A" and 36 nurses and 190 nursing technicians at hospital "B".The researcher's convenience, the availability of the women to participate in the research and at least six months on the job were used as inclusion criteria.
Each nurse and nursing technician received a selfadministered questionnaire in their place of work, including instructions for completion.A deadline was set to return the completed questionnaire, with a maximum of three returns to collect all instruments.Data collection at each hospital was closed off after two weeks.

Study variables and data collection instrument
The study variables and causal relations among them have been detailed in Figure 1.The respondent scored each adjective twice: first considering the perception of the professional category (self-perception), followed by society's perception of that professional category (hetero-perception).
To strong bonds with the institution; feeling of institutional support; sufficient knowledge about the institution (7) .To assess agreement with or disagreement from each of the situations, a five-point Likert scale was used.
To assess the questions about experiences of pleasure and suffering, the Scale of Indicators of Pleasure and Suffering at Work -EIPST was used.
This scale is part of an inventory about work and risk of illness (ITRA), previously validated, built based on the theoretical framework of the psychodynamics of work and comprising four independent scales (8) .The EIPST consists of four dimensions, two for pleasure and two for suffering: the first indicates professional accomplishment and freedom of expression; the second emotional exhaustion and lack of recognition.The frequency of positive and negative experiences was assessed using a five-point Likert scale (1= never; 5= always).To maintain a valid and reliable instrument, the scale was reduced by applying Spearman-Brown's formula, maintaining a number of items in each of the four dimensions to permit a minimum reliability of 0.700 (Cronbach's alpha).This means that the employed instrument is a short version of the original (8) . The

Data recording and analysis
The participants' answers were recorded in a  (9) .Initially, cases with a loss of two standard deviations above the group average were excluded, and more than two standard deviations from the mean number of data missing from the base.Nine cases were excluded from each hospital database.These missing data were replaced through the so-called 'expectationmaximization' procedure, with a view to finding that maximum likelihood of parameters, resulting in 531 complete cases in the research sample.
The identification of extreme cases as those beyond the limits of ±2.58 from the mean revealed 609 extreme answers (0.96% of the base).The skewness and kurtosis parameters in each group revealed significant (Z value beyond limits of ±1.96) and expressive (measures beyond the limits of ±1) discrepancies for practically all variables (10) .Similarly,

Results
In the total sample at the two hospitals, 48.2% of the women had no children and 44,2% had one or two children; 44.2% were single and 42.9% married.
Concerning their time on the job at the hospital, 26.4% had worked there less than one year, 49.9% between one and five years and 18.9% more than ten years; 19% were baccalaureate nurses and 81% nursing technicians; 73.3% worked day shifts.The mean age was 36.77years (standard deviation = 9.97).
Table 1 indicates the results of the exploratory factor analysis: Vieira A, Alves M, Monteiro PRR, Garcia FC.

Table 1 -(continuation)
Kaiser Meyer Olkin (KMO) Measure of sampling adequacy = 0.918.Extracted variance = 67.9.Favorable measures for the application of EFA (10)  For the scale of experiences of pleasure and suffering at work (8) , after the exclusion of four items (freedom to use my creativity, insecurity, fear and uselessness) from those initially selected for the instrument (8) , due to low communality and/or crossed loadings, a satisfactory factorial solution was reached, as demonstrated in Table 2.
Favorable measures for the application of EFA (10)  to the actual score of the latent variable, that is, the variance percentage free from random measurement errors.Values typically higher than 0.700 are accepted but, for scales under construction, values up to 0.600 can be included (9) .These measures show that the scales possess acceptable properties, as most values are superior to 0.800.The AVErepresents the mean reliability of the indicators and should be higher than 0.400, which was the case for all of the scales.

Indicators
Then, the validity of the measures was assessed through the construct validity procedure (9) and the structural equations modeling approach was applied for this purpose (11) , using the partial least squares estimation method, which is the most appropriate given the relation between sample and model size and the presence of deviations from normality (12) .
Construct validity consists of three components: convergent validity; discriminant validity and nomological validity.In convergent validity, the intent is to identify whether all indicators originate in the same factor, which means the evaluation of whether the factor loadings of the indicators are significant at 5% or 1% one-tailed (12) .This criterion revealed that practically all indicators complied with the proposed requisite with good reliability, given that the squared coefficient of the standardized factor loadings was higher than 0.4 in most cases.To measure discriminant validity (12) , the Average Vieira A, Alves M, Monteiro PRR, Garcia FC.Vieira A, Alves M, Monteiro PRR, Garcia FC.
possesses low intellectual complexity tend to experience greater suffering and less pleasure.
The dimension 'ethics' only affected the experience of suffering at work, with positive relations.This demonstrates that, as the self-perception of high ethical standards inherent in the profession increases, while society does not recognize these standards, professional suffering increases.
The variable organizational identification, in turn, exerted a strong positive influence on experiences of pleasure and a negative influence on experiences of suffering.

Conclusion
The general aim in this paper was to analyze the influence of self-perception and hetero-perception on the organizational identification and experiences of pleasure and suffering of female nursing team members at two general hospitals in Belo Horizonte/MG.In conclusion, the organizational identification variable exerts a considerable positive impact on the experiences of pleasure and a negative impact on the experiences of suffering.On the other hand, no causal relation could be established between self-and hetero-perception and organizational identification.In other words, it cannot be affirmed that, the more positive the self-and heteroperception and the greater the agreement between both, the higher the individuals' degree of identification with the organization will be.The present results cannot be compared either with others, as no similar studies have been found in the scientific literature.
The methodological limitation of this research is related to the sample size, which could be expanded to permit conceptually more robust statistical analyses like structural equations modeling, using the analysis of the identification process with organizations and social recognition for one's work can minimize the suffering and project the pleasure.A survey in health journals showed no publications that linked organizational identification with experiences of pleasure and suffering at work, nor scales to assess self-and hetero-perception.Also, most publications on identity and perception of professional image prioritize the use of qualitative methods.With a view to completing this knowledge gap, this paper proposes a quantitative model to analyze the behavior of the study variables.Method This descriptive, correlational and cross-sectional research with a quantitative approach involved 511 nursing team members at two general hospitals in Belo Horizonte/MG.Hospital "A" is affiliated with the municipal health network, exclusively functions within the Unified Health Network (SUS) and offers 409 beds, with a staff of 2,553 professionals.Hospital "B" is nonprofit, delivers care to patients forwarded by health insurances and the SUS and offers 360 beds, with a 1,500-member staff.Data were collected in November and December 2012.The participants' anonymity was guaranteed in compliance with the ethical principles of National Health Council Resolution 196/96.Approval for the research was obtained from the Research Ethics Committee at the hospitals: "A" (protocol 385866/2010 on 12/09/2012) and "B" (protocol 370/11 on 05/26/2011) and the study was conducted in compliance with ethical requirements.All participants signed the informed consent form and data confidentiality was guaranteed.

Figure 1 - 1 (
Figure 1 -Research Model These cases were replaced by less offensive values within the scale limits.The identification of multivariate extreme cases according to the Mahalanobis distance (measure of the standard distance of each point from the centroid in a multivariate correlated plan through which aberrant patterns can be identified and analyzed), in comparison with a chi-squared statistic with an 0.1% significance level and 119 degrees of freedom revealed four extreme cases for one hospital and 15 for the other.With a view to maintaining an appropriate sample size to test the model and assess the possible influence of the cases in the analysis, tests were accomplished with and without extreme cases in the research.The differences between the results were minimal and not perceptible, causing no changes in the study conclusions.Next, compliance with the normality assumption was assessed through the Z significance tests of kurtosis and skewness, as well as the Kolmogorov-Smirnov test.
global significance tests demonstrated noticeable deviations from normality in the above mentioned groups, denoting a discrepancy that encourages the use of analytical methods more robust against the violation of the normality assumption, in this case the Partial Least Squares (PLS) method.In addition, the linearity and multicollinearity assumptions were assessed through the Variance Inflation Factor (VIF), the scatterplots (for 100 pairs of variables) and the sample correlation matrix.Evidences demonstrated that the redundancy level (VIF below 10 and correlations between ±0.9) and the previous mentioned assumptions were achieved.Next, the quality of the scales was assessed through the Exploratory Factor Analysis (extraction through principal axis analysis and direct oblimin rotation).Initially, the self-perception responses were used for the test, followed by the hetero-perception responses.The results of this assessment showed that some indicators did not fall into their respective factors, although the hetero-perception scale revealed good conceptual coherence.Given the goal of using the difference between the perceptions and their influence on the identification and on experiences of pleasure and suffering, in the analysis, the difference between these indicators was applied to identify the underlying factors.
covariance structures.For future research, qualitative analyses are also suggested, with a view enhancing the interpretation of the study variables' behavior.The explanatory percentages found reveal fertile ground to assess the influence of other individual (personality, values, life experience) and contextual factors (quality of life at work, infrastructure, remuneration) on organizational identification as well as on experiences of pleasure and suffering at work in future studies.

Table 4 -
Structural relations tested in complete sampleVariance Extracted of the indicators was compared with the squared coefficient of the correlations between the pairs of constructs.All scales demonstrated discriminant validity according to these criteria.The results of the preceding analysis demonstrated that the constructs possess discriminant validity.Finally, the general research model was applied to the presented data.Using the bootstrap method, it was verified which relations between self-and hetero-perception were statistically significant at 5% of one-tailed significance (Table4).
*Weight: weight of structural relation assessed in the model †Error: standard error of weight ‡T-value: statistics of significance test of structural relation (if higher than 1.96 p < 0.05 two-tailed, if higher than 1.645 p < 0.05 one-tailed)