Integrating Gender Equity into Psychosocial Risk Assessment: Development and Pilot Validation of a Perceived Gender Disparity Subscale Within the OPRA Framework

Abstract

Background: Gender equity is an increasingly relevant dimension in psychosocial risk assessment within organisational settings. This pilot study examines the integration of a Perceived Gender Disparity subscale into the Organisational and Psychosocial Risk Assessment (OPRA) inventory in an Italian manufacturing sample (N = 65). Methods: The five-item subscale, grounded in the literature on workplace gender discrimination and harassment, underwent reliability analysis (Cronbach’s α = 0.706), bivariate correlations (Pearson’s r), Exploratory Factor Analysis (EFA), non-parametric sensitivity analysis (Spearman’s ρ), and exploratory gender comparisons (exact Mann–Whitney U tests). Results: Results indicate statistically significant associations with seven of nine OPRA subscales: a large correlation with Role (r = 0.508, p < 0.001), medium-sized associations with Organisational Culture, Workload, Quality of Relationships, and Career Development (|r| = 0.331–0.427, p ≤ 0.007), and small-to-medium associations with Environment and Safety and Autonomy (|r| = 0.268–0.275, p < 0.05). No significant associations emerged for Working Hours or Work–Life Balance. Spearman’s correlations confirmed and strengthened all significant associations. Exploratory gender comparisons revealed medium-sized effects in the expected direction on the items assessing gender-specific disadvantage. EFA revealed a unidimensional structure with adequate loadings (λ = 0.475–0.785), accounting for 41.90% of the total variance. Conclusions: This study constitutes an initial psychometric evaluation rather than a definitive validation. Despite the limitations of a single-site, male-dominated sample, the findings provide preliminary support for including perceived gender disparity as a source of organisational risk within the OPRA framework. Larger and more diverse samples are required to confirm the factor structure and to establish the generalizability of the findings.

1. Introduction

Attention to the gender dimension in occupational health has accelerated significantly over the past decades. Transformations in the labour market—rising female participation, the evolution of organisational structures, and the emergence of inclusivity demands—have rendered any gender-neutral approach to psychosocial risk that fails to account for structural gender-based asymmetries insufficient. Although women represent approximately 70% of the workforce in the health and social care sector, they hold only 25% of senior leadership positions (Smith and Sinkford 2022), and in many settings, they face a dual burden of work and domestic responsibilities within environments that tend to underestimate stress-related risk (European Agency for Safety and Health at Work 2023).

A particularly robust indicator of such asymmetries is the glass ceiling phenomenon: the set of invisible barriers that hinder the advancement of qualified female workers toward top-level positions. The sustaining mechanisms identified by Babic and Hansez (2021) are the lack of mentoring, the absence of objective evaluation criteria, and an organisational culture that marginalizes women on the basis of maternity-related stereotypes. Female workers who perceive such barriers experience a marked Work-to-Family Conflict, which mediates the effects of the glass ceiling on increased job strain and reduced job engagement. More broadly, organisations characterized by a predominantly masculine culture tend to associate management with qualities such as authority and aggressive assertiveness, penalizing women through a likeability penalty whenever they adopt such behavioural styles (Smith and Sinkford 2022).

A comprehensive understanding of organisational inequalities further requires recourse to the concept of intersectionality: systems of disadvantage based on gender, ethnicity, and sexual orientation produce dynamics that are not reducible to their individual components. Workers belonging to the LGBTQIA+ community report systematically lower levels of occupational health, which can be interpreted in light of the Minority Stress Theory, according to which chronic stress arises from prolonged exposure to stigma, prejudice, and microaggressions in the workplace (Oliveira et al. 2025). This body of evidence converges in identifying organisational culture as the primary mediator between formal diversity management policies and worker-perceived outcomes (Alonso Gallo and Gutiérrez López 2023).

It is within this framework that the OPRA (Organisational and Psychosocial Risk Assessment)—a validated questionnaire for the assessment of psychosocial risk and organisational well-being—is situated (Magnavita 2008; Andrisano Ruggieri et al. 2016a, 2016b, 2017, 2018, 2020, 2025). In its risk sources section, the OPRA maps nine dimensions of the work experience—including interpersonal relationships, workload, autonomy, organisational culture, and career development—thus providing a diagnostic framework applicable to both occupational health interventions and human resource development policies. However, in its original formulation, the instrument does not incorporate an explicit measure of perceived gender disparity. Given the growing body of evidence suggesting that gender should be treated not as a control variable but as a fundamental organizer of the psychosocial risk experience (Fida et al. 2023), this diagnostic gap warrants attention.

The present pilot study aims to test the integration of a Perceived Gender Disparity subscale into the OPRA inventory by examining its factor structure and its associations with the risk dimensions already covered by the instrument in an Italian manufacturing sample. Consistent with this exploratory purpose, the study is conceived as an initial psychometric evaluation of the proposed subscale, intended to provide a foundation for subsequent validation work on larger and more heterogeneous samples, and not as a definitive validation of the instrument.

2. Materials and Methods

The study was conducted at Officine Meccaniche Irpine (OMI), an Italian metalworking company. A total of 65 employees participated in the survey, representing 70.65% of the company workforce (N = 92). All participants provided informed consent prior to taking part in the study. Participation was voluntary and anonymous, and respondents were informed of their right to withdraw at any time without consequence. The sample was predominantly male (83.1%; n = 54), with a female proportion of 16.9% (n = 11). Company records indicate that the workforce comprised 81 male and 11 female employees. The participation rate was 54/81 (66.7%) among men and 11/11 (100%) among women, reflecting the gender segregation typically observed in the manufacturing sector. The mean age was 40.7 years (SD = 11.4; range: 20–63); 89.2% of respondents held permanent employment contracts. The occupational composition was characterized by a prevalence of blue-collar workers (49.2%), followed by white-collar employees (44.6%), middle managers (4.6%), and one senior manager (1.5%). Data were collected anonymously through the digital administration of the extended version of the OPRA questionnaire.

It should be noted that the strong male prevalence in the sample constitutes not only a limitation in terms of statistical power, but also a constraint on the ecological validity of the subscale: the data predominantly reflect male perceptions of gender disparity, which may differ qualitatively from the lived experience of female workers. This aspect will need to be addressed in future studies employing samples with a more balanced gender composition.

The Perceived Gender Disparity subscale was developed to extend the original OPRA framework through the inclusion of five items calibrated to the Italian organisational context. The conceptual and methodological development of the subscale drew upon established instruments designed to assess gender equity and inclusion in organisational environments, particularly the Survey on Gender, Equity and Inclusion developed by Theofanos and colleagues at the National Institute of Standards and Technology (Theofanos et al. 2021). This instrument served as a key methodological reference, particularly regarding the operationalization of perceived inequities, organisational climate, and subtle forms of gender-based exclusion.

Item development followed a structured three-phase process. First, a targeted literature review was conducted to identify the principal dimensions of gender-based inequality within organisational settings, with particular attention to constructs such as the glass ceiling (Babic and Hansez 2021), sexist harassment (Tricco et al. 2024), and gender microaggressions (Oliveira et al. 2025). Second, the research team formulated a preliminary pool of items designed to capture perceived inequities in opportunities, recognition, and organisational treatment. Particular care was taken to ensure conceptual alignment and linguistic consistency with the structure and response format of existing OPRA subscales.

Third, the preliminary items underwent a structured content validity procedure involving a panel of five experts in work and organisational psychology with experience in psychosocial risk assessment and gender studies. Experts independently evaluated each item in terms of clarity, relevance, and representativeness of the construct. Based on their feedback, minor wording revisions were introduced to improve semantic precision and contextual appropriateness. One preliminary item was removed due to conceptual overlap with existing OPRA dimensions, resulting in the final five-item subscale. To further ensure comprehensibility and ecological validity, the revised items were pilot-tested with a small group of workers drawn from the same organisational context. An informal cognitive debriefing procedure indicated adequate clarity and interpretability, supporting the adoption of the final item set in the present study. This multi-step development process was designed to ensure both theoretical grounding and contextual appropriateness of the subscale within the OPRA diagnostic framework.

The five items assess perceived gender discrimination and disparity across five domains: (1) the perceived need to work harder in order to receive equivalent recognition; (2) the perception that one’s professional judgement is questioned on the basis of gender; (3) loss of work opportunities due to gender; (4) direct experience of sexist harassment or abuse; and (5) awareness of harassment episodes experienced by colleagues. It should be noted that the first three items capture first-person perceived discrimination, whereas items 4 and 5 focus on direct exposure and vicarious exposure to harassment, respectively. This distinction is relevant insofar as vicarious exposure, while correlated with the perception of a hostile organisational climate, constitutes a partially distinct construct from personal discrimination. The factor structure reported below allows for an empirical examination of the internal coherence of this item set.

Responses are anchored to the same five-point Likert frequency scale used across the remaining OPRA sections (1 = Never; 2 = Rarely; 3 = Sometimes; 4 = Often; 5 = Always). Higher scores on the subscale indicate a stronger perception of gender disparity, whereas on the nine original OPRA subscales, higher scores indicate greater perceived severity for risk dimensions (Role, Workload, Quality of Relationships, Career Development, Working Hours, Work–Life Balance) and, conversely, a more favourable perceived condition for protective dimensions (Autonomy, Environment and Safety, Organisational Culture and Structure). This asymmetry in scoring direction is relevant to the interpretation of the correlations reported in the following section. The internal consistency of the subscale in the present sample proved acceptable (Cronbach’s α = 0.706), a value that meets the conventional threshold of 0.70 generally adopted for scales at an early stage of development (Nunnally and Bernstein 1994), although modest for a scale comprising only five items—a condition under which alpha tends to be mechanically deflated by the low number of items (Cortina 1993). To assess internal consistency at the item level, corrected item-total correlations were computed for each of the five items comprising the Perceived Gender Disparity scale. This procedure estimates the degree of association between an individual item and the composite score derived from the remaining items, with the item itself excluded from the calculation in order to avoid spurious self-correlations. In psychometric research, the conventional threshold for item adequacy is set at r = 0.30, with values above this threshold indicating that an item contributes meaningfully to the measurement of the same underlying construct (Nunnally and Bernstein 1994).

3. Results

Means and standard deviations for all subscales are reported in

Table 1. Descriptive statistics for the OPRA subscales and the Perceived Gender Disparity subscale (N = 65).

Subscale	M	SD
Work Environment and Safety	3.98	0.73
Autonomy	3.49	0.83
Organisational Culture	3.22	0.74
Career Development	3.14	0.44
Workload	2.94	0.77
Work–Life Balance	2.83	0.48
Quality of Relationships	2.33	0.35
Role	2.31	0.72
Working Hours	2.02	0.55
Perceived Gender Disparity	1.45	0.50

Note: Scores given on a 1–5 scale. For the original OPRA subscales, higher scores indicate greater perceived severity (Role, Workload, Quality of Relationships, Career Development, Working Hours, Work–Life Balance) or a more favorable perceived condition (Autonomy, Environment and Safety, Organisational Culture and Structure). For the Perceived Gender Disparity subscale, higher scores indicate a stronger perception of gender disparity.

. The Perceived Gender Disparity subscale yielded the lowest mean of the entire set (M = 1.45, SD = 0.50), indicating that perceived gender disparity in the present sample was at low levels overall—a finding consistent with the strong male predominance in the manufacturing context, where gender-related pressure may be less salient for the majority of respondents. The protective dimensions of the OPRA (Environment and Safety, Autonomy) showed the highest mean values, whereas the risk dimensions (Role, Working Hours, Quality of Relationships) fell within the low-to-moderate range.

The results of the analysis are reported in

Table 2. Bivariate correlations between the Perceived Gender Disparity subscale and the OPRA risk dimensions.

Variable	1	2	3	4	5	6	7	8	9	10
1. Perceived Gender Disparity	—
2. Organisational Culture	−0.43 ***	—
3. Career Development	0.33 **	−0.24	—
4. Role	0.51 ***	−0.69 ***	0.25 *	—
5. Autonomy	−0.27 *	0.67 ***	−0.01	−0.60 ***	—
6. Work–Life Balance	0.09	0.19	0.19	−0.06	0.22	—
7. Work Environment and Safety	−0.28 *	0.71 ***	−0.11	−0.55 ***	0.63 ***	0.21	—
8. Workload	0.40 **	−0.56 ***	0.35 **	0.45 ***	−0.27 *	0.17	−0.44 ***	—
9. Working Hours	0.15	−0.38 **	0.14	0.36 **	−0.35 **	−0.01	−0.44 ***	0.35 **	—
10. Quality of Relationships	0.39 **	−0.15	0.31 *	0.36 **	−0.20	0.24	−0.02	0.33 **	0.11	—

Note: N = 65. The dash (—) indicates the correlation of a variable with itself. * p < 0.05. ** p < 0.01. *** p < 0.001.

. The Perceived Gender Disparity subscale showed statistically significant associations with seven of the nine OPRA subscales.

Correlation analyses revealed a pattern of statistically significant associations between the Perceived Gender Disparity subscale and the OPRA dimensions. The largest correlation was found with the Role subscale (r = 0.508, p < 0.001), classifiable as a large correlation according to Cohen (2013). Positive correlations of medium magnitude emerged for Workload (r = 0.396, p = 0.001), Quality of Relationships (r = 0.392, p = 0.001), and Career Development (r = 0.331, p = 0.007).

The negative correlations reflect the scoring direction asymmetry inherent in certain OPRA subscales, where higher values indicate a more favourably perceived condition: the largest negative coefficient was recorded for Organisational Culture and Structure (r = −0.427, p < 0.001), followed by Environment and Safety (r = −0.275, p = 0.026) and Autonomy (r = −0.268, p = 0.031). The Working Hours (r = 0.145, p = 0.248) and Work–Life Balance (r = 0.087, p = 0.489) subscales did not show significant covariations—a finding plausibly attributable both to the limited statistical power of the sample and to the characteristics of the organisational context: in a manufacturing organisation with structured shift scheduling, working hours constitute a variable with restricted internal variability, which is perceived homogeneously regardless of gender.

To verify the robustness of the findings and to account for the distributional characteristics of the data, all variables under examination were subjected to the Shapiro–Wilk test of normality. As anticipated, owing to the floor effect discussed previously, the distribution of the focal subscale Perceived Gender Disparity deviated significantly from normality (W = 0.838, p < 0.001), as did several other dimensions of the inventory. To address this distributional skewness, which risks attenuating linear estimates mathematically, the associations between constructs were recalculated using Spearman’s non-parametric rank correlation coefficient (ρ). When the associations were re-estimated on a monotonic basis, without assuming linearity, they proved markedly stronger: in particular, the association with the Role dimension increased from a moderate to a substantially large effect size (ρ = 0.607, p < 0.001), as did the negative relationship with Organisational Culture and Structure (ρ = −0.516, p < 0.001). Comparable increments were observed for Quality of Relationships (ρ = 0.440, p < 0.001) and Workload (ρ = 0.420, p < 0.001). Overall, this sensitivity analysis indicates that the associations between perceived gender disparity and psychosocial risk dimensions are likely stronger than the parametric estimates initially suggested, once distributional constraints are considered.

In line with the procedure described in the Methods section, all corrected item-total correlations exceeded the conventional criterion of 0.30, with values of 0.435 for Item 1, 0.690 for Item 2, 0.553 for Item 3, 0.440 for Item 4, and 0.494 for Item 5, indicating that each item contributes adequately to the measurement of the underlying construct.

To assess the construct validity of the new subscale, an Exploratory Factor Analysis (EFA) was conducted, following established methodological recommendations (Fabrigar et al. 1999; Costello and Osborne 2005). The suitability of the correlation matrix for factor extraction was first evaluated by computing the Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy and Bartlett’s test of sphericity. The KMO value of 0.655 falls below the conventional threshold of 0.70 (Kaiser 1974), qualifying as “mediocre” according to the taxonomy proposed by Hutcheson and Sofroniou (1999); however, it exceeds the minimum threshold of 0.50 below which factor analysis would be considered inappropriate. This value, together with the observation-to-item ratio at the lower bound of acceptability (13:1), warrants caution in interpreting the obtained factor solution. Bartlett’s test of sphericity was significant (χ² = 112.22; df = 10; p < 0.001), confirming that the correlation matrix differs significantly from the identity matrix.

The single-factor extraction yielded a solution accounting for 41.90% of the total variance. While this value falls below the 50% threshold sometimes recommended for unidimensional solutions, it is consistent with the partial conceptual heterogeneity of the item set and with the sample size limitations discussed above. All five items exceeded the inclusion threshold of λ = 0.30. Factor loadings are reported in

Table 3. Factor loadings of the Perceived Gender Disparity subscale (N = 65).

Item	λ
2. Professional judgement questioned on grounds of gender bias	0.785
5. Awareness of harassment incidents in the workplace	0.672
4. Direct experience of sexist harassment	0.642
3. Perceived loss of professional opportunities	0.624
1. Necessity of greater effort for equivalent recognition	0.475

Note: Extraction method: Principal Axis Factoring. Single-factor extraction.

. In view of the sample size, the factor solution reported here should be regarded as provisional and as requiring confirmation in independent and larger samples.

The highest factor loadings were found for the perception that one’s professional judgement is questioned on the basis of gender (λ = 0.785) and awareness of harassment episodes in the workplace (λ = 0.672), followed by direct experience of sexist harassment (λ = 0.642) and perceived loss of professional opportunities (λ = 0.624). The item concerning the need to exert greater effort in order to obtain equivalent recognition yielded the lowest loading (λ = 0.475), while still exceeding the acceptability threshold. Notably, the items related to direct and vicarious exposure to harassment (items 4 and 5) exhibited loadings comparable to those of the perceived discrimination items (items 2 and 3), providing empirical support for the internal coherence of the scale despite the partial conceptual heterogeneity noted in the subscale development section.

In response to the exploratory aim of examining response patterns across gender groups, male (n = 54) and female (n = 11) participants were compared based on scores for the Perceived Gender Disparity subscale and its constituent items using exact Mann–Whitney U tests, using the rank-biserial correlation (r) as the effect size estimate. Exploratory comparisons were also extended to the four OPRA dimensions most directly tied to the structural conditions of work, namely Work Environment and Safety, Autonomy, Workload, and Working Hours. Given the small size of the female subsample, and considering that no correction for multiple comparisons was applied, these analyses should be regarded as strictly exploratory. Results are reported in

Table 4. Exploratory gender comparisons on the Perceived Gender Disparity subscale, its constituent items, and four OPRA dimensions (N = 65).

Variable	Men (n = 54) M (SD)	Women (n = 11) M (SD)	U	Exact p	r
Perceived Gender Disparity (total)	1.40 (0.44)	1.71 (0.70)	355.5	0.319	0.20
Item 1 (greater effort)	2.22 (1.11)	2.18 (1.25)	288.0	0.883	−0.03
Item 2 (judgement questioned)	1.33 (0.67)	1.91 (0.94)	399.0	0.076	0.34
Item 3 (lost opportunities)	1.31 (0.70)	1.91 (1.04)	401.0	0.070	0.35
Item 4 (direct harassment)	1.07 (0.38)	1.27 (0.47)	364.0	0.248	0.23
Item 5 (awareness of harassment)	1.06 (0.30)	1.27 (0.47)	365.5	0.241	0.23
Work Environment and Safety	3.95 (0.75)	4.12 (0.70)	341.0	0.452	0.15
Autonomy	3.44 (0.83)	3.73 (0.85)	354.5	0.328	0.19
Workload	2.94 (0.79)	2.93 (0.66)	297.5	1.000	0.00
Working Hours	2.10 (0.55)	1.61 (0.38)	137.0	0.004	−0.54

Note: U = exact Mann–Whitney U statistic; r = rank-biserial correlation, with positive values indicating higher scores among female participants. No correction for multiple comparisons was applied.

.

Female participants reported descriptively higher scores on the total subscale (Mdn = 1.60 vs. 1.20; U = 355.5, exact p = 0.319, r = 0.20), a difference in the expected direction that did not reach statistical significance. At the item level, medium-sized effects emerged for the perception that one’s professional judgement is questioned on the basis of gender (U = 399.0, exact p = 0.076, r = 0.34) and for the perceived loss of professional opportunities (U = 401.0, exact p = 0.070, r = 0.35): 54.5% of female participants, compared with approximately 21% of male participants, reported at least occasional experiences in both domains. Endorsement of direct and vicarious exposure to harassment was likewise markedly higher among women (27.3% vs. 3.7% for both items). Notably, the only item showing no gender differentiation was the perceived need to work harder for equivalent recognition (r = −0.03), which is also the item yielding the lowest factor loading (λ = 0.475), suggesting that male respondents may interpret this item in generic meritocratic rather than gender-specific terms. Among the four OPRA dimensions examined, the only significant difference concerned Working Hours (U = 137.0, exact p = 0.004, r = −0.54), with lower scores among women, a result largely attributable to the occupational composition of the two groups: ten of the eleven female participants worked a regular daytime schedule, whereas approximately one third of male participants were shift workers. This finding is consistent with the interpretation, advanced above, that in the present organisational context, working hours are structured primarily by shift scheduling rather than by gender.

4. Discussion

The results reveal a coherent and theoretically interpretable correlational profile, providing preliminary empirical support for the integration of the Perceived Gender Disparity subscale into the OPRA diagnostic set. The most salient finding is the large correlation with the Role dimension: the association between perceived gender disparity and difficulties in work role definition suggests that gender disparity acts as a factor that penetrates the core of professional identity, consistent with Social Role Theory (Eagly 2013), according to which stereotyped role expectations exert direct effects on perceived self-efficacy and on the appraisal of one’s own competence within the organisational context.

The association with Workload deserves specific consideration: in a strongly male-dominated context such as the manufacturing setting analysed (83.1% men), the positive correlation may reflect not so much differences between genders, but rather the way in which organisational culture, with its implicit performance expectations, affects the perception of work pressure across genders—a dynamic akin to the dual quantitative and emotional burden documented in the literature for predominantly masculine organisational cultures.

As documented by Fida et al. (2023), in settings with lower occupational segregation, differences in the perception of stressors tend to diminish while persisting in more deeply rooted cultural dimensions. The correlation with Quality of Relationships reflects a higher frequency of hostile behaviours and dynamics of exclusion from informal communication networks, conceptualizable as gender microaggressions (Oliveira et al. 2025); the correlation with Career Development, in turn, aligns with the literature on the glass ceiling (Babic and Hansez 2021), indicating that the perception of disparity translates into a more critical appraisal of available advancement opportunities.

The negative correlations with Autonomy and Environment and Safety, which were of small-to-medium magnitude, are consistent with the literature on gendered power asymmetries: those who perceive greater inequalities tend to evaluate the conditions of control over their own work less favourably as well as the safety of the physical environment, suggesting that the perception of disparity spreads across multiple domains of the work experience. The negative association with Organisational Culture and Structure represents the most relevant finding from an intervention standpoint: workers who perceive gender inequalities tend to judge the entire communicative and participatory architecture of the organisation as less transparent and less responsive to individual needs, confirming perceived gender disparity as a cross-cutting indicator of organisational culture quality. This is corroborated by the inclusive climate models proposed in the literature (Alonso Gallo and Gutiérrez López 2023), which identify organisational culture as the primary mediator between formal diversity management policies and worker-perceived outcomes.

The exploratory gender comparisons enrich this picture in a manner consistent with the intended function of the subscale. Although the overall difference between men and women did not reach statistical significance, the items addressing gender-specific disadvantage, namely the questioning of professional judgement and the loss of professional opportunities, showed medium-sized effects in the expected direction, and the endorsement of direct and vicarious exposure to harassment was several times higher among female participants. The convergence between the absence of gender differentiation on the item concerning greater effort for equivalent recognition and its comparatively low factor loading further suggests that this item may partly capture generic meritocratic concerns rather than gender-specific disadvantage, an aspect worth examining in future refinements of the subscale. Taken together, these patterns support the sensitivity of the instrument to the experiences of the group most directly exposed to gender-based disadvantage; however, we reiterate that inferences drawn from a subsample of eleven women must remain provisional.

It is appropriate to make explicit the rationale underlying the denomination adopted for the subscale. The items are oriented toward the detection of experiences of discrimination, perceived disadvantage, and harassing behaviours within organisational settings; they do not directly measure gender equity in a positive sense, which would instead entail the assessment of inclusive organisational policies, equal career opportunities, and the transparency and fairness of evaluation and advancement criteria. For this reason, the denomination “Perceived Gender Disparity” was preferred over “Gender Equity”, as it is more consistent with the actual content of the items and more likely to avoid a potential conceptual ambiguity between the measurement of discrimination or disadvantaged conditions and the measurement of equity as a positive organisational condition.

At the same time, it should be noted that data collection was carried out within an organisation that, over the past two years, has implemented a structured programme for the promotion of gender equity. This programme extends beyond internal company policies to encompass the surrounding community in which the organisation operates, through initiatives and interventions directed at local schools, municipal administrations, and families in the area. This organisational context represents a relevant element for the interpretation of the results as it suggests that the perception of potential disparities is situated within an environment that has already undertaken an explicit path toward the promotion of gender equity.

5. Conclusions and Limitations

The present pilot study has limitations that circumscribe its scope and that must be clearly acknowledged. The principal limitation concerns the size and composition of the sample. With only 65 participants and a small female component (n = 11), statistical power is limited, with a non-negligible risk of Type II error for the weaker coefficients. The non-significant correlations with Working Hours and Work–Life Balance may partly reflect this constraint, in addition to the organisational specificities already discussed. More critically, the question of ecological validity arises: the sample is almost entirely male, meaning that the data predominantly reflect male perceptions of gender disparity. Such perceptions may differ qualitatively from those of female workers, who constitute the group most directly exposed to discrimination. The marked floor effect observed (M = 1.45 on a 1–5 scale) is consistent with this interpretation: in a sample where the majority is not the primary target of gender discrimination, responses tend to cluster at the lower end of the scale. It is therefore not possible to assume that the correlational patterns would replicate in samples with a more balanced gender composition. For the same reason, the exploratory gender comparisons reported above, which were conducted without correction for multiple testing, should be interpreted as preliminary indications rather than as stable estimates of group differences.

The study design is a single-site study: the data originate from a single company in the metalworking sector, with its own cultural, organisational, and shift-scheduling specificities. Generalizability to different work settings (service sector, health and social care, public administration) remains entirely to be established.

The observation-to-item ratio (13:1) and the KMO value (0.655) both fall at the lower bound of the thresholds recommended for EFA, which warrants caution in interpreting the factor structure. The obtained loadings are encouraging, but the factor structure reported here must be regarded as provisional: its confirmation in larger and more sectorially diverse samples is a necessary condition before the subscale can be considered psychometrically established.

Moreover, the cross-sectional nature of the research design does not allow for the identification of causal relationships among the dimensions examined.

Finally, the item development process, although grounded in the literature and subjected to expert review, relied on an informal cognitive debriefing conducted with workers drawn from the same organisational context, and did not include a formal cognitive interviewing phase or piloting on an independent sample, steps that are recommended as best practices in psychometric scale development (DeVellis and Thorpe 2022).

The results of the present pilot study indicate that the Perceived Gender Disparity subscale integrates organically into the OPRA diagnostic framework, offering a non-redundant informational contribution that is theoretically consistent with the risk dimensions already covered by the instrument. The correlational pattern obtained confirms that the perception of gender inequality systemically articulates with the principal sources of organisational risk—from role-related difficulties to the deterioration of interpersonal relationships, through to a negative appraisal of organisational culture—in line with the proposal by Fida et al. (2023) to treat gender as a fundamental organizer of the psychosocial risk experience, rather than as a mere control variable.

From an applied perspective, the systematic adoption of the subscale would address a significant diagnostic gap: at a time when organisations are being increasingly called upon to respond to growing demands for equity and inclusivity, having access to measurement instruments sensitive to these dimensions represents a strategic lever for the planning of effective corrective interventions.

Future research should prioritize three lines of inquiry. First, replication in larger (N ≥ 200) and sectorially diversified samples with a balanced gender composition, in order to verify the stability of the correlational pattern and to conduct a Confirmatory Factor Analysis (CFA) formally testing the unidimensionality hypothesis (Kline 2023). Second, the examination of criterion validity through correlations with health outcomes, well-being, and job satisfaction, consistent with the original OPRA validation protocol (Magnavita 2008). Third, the investigation of differences in response patterns between men and women (and, where sample size permits, measurement invariance analysis) in order to ascertain whether the subscale operates equivalently across groups—a necessary condition for its reliable diagnostic use in organisational settings with heterogeneous gender compositions. Until such evidence becomes available, the Perceived Gender Disparity subscale should be regarded as a promising but provisional measure, and the present findings as an initial psychometric evaluation that offers a foundation for further research rather than conclusive validation.