Methodology

This database documents professional experiences of women and non-binary people in the creative industries of the German-speaking region (DACH: Germany, Austria, Switzerland). Throughout this page and the entire platform, after the initial full reference, we use the term affected persons as an abbreviated designation.

The choice of this terminology is grounded in the current scientific and legal-political discourse on gender-based disadvantage. The Deutscher Frauenrat (German Women's Council), the largest umbrella organization of German women's associations with over 60 member organizations, uses comparable plain-language formulations and explicitly includes trans women (Deutscher Frauenrat, 2023). All three DACH countries now legally recognize non-binary gender identities: Germany through the Self-Determination Act (SBGG, 2024), Austria through the Constitutional Court decision (VfGH, 2018), and Switzerland through increasing administrative recognition at the cantonal level. The acronym FLINTA* is deliberately avoided, as it is not generally understood without explanation and has so far gained little traction in academic literature (Lemke, 2023).

Intersectionality

This project understands gender-based disadvantage as part of overlapping structures of discrimination. Kimberle Crenshaw demonstrated that anti-discrimination jurisprudence, by treating gender and race separately, rendered the experiences of multiply marginalized individuals invisible (Crenshaw, 1989). In further developing her concept, Crenshaw (1991) distinguished between structural, political, and representational intersectionality. Gender interacts with race, disability, age, and other identity characteristics. The database does not currently capture these intersections systematically but acknowledges their significance for understanding professional disadvantage.

Scope

The database captures experiences from the creative industries — Audio, Film, and Games — in the German-speaking region (Germany, Austria, Switzerland). The scope is defined by the self-identification of contributing individuals, not by external attribution — in accordance with the self-determination principle of the SBGG (2024). An expansion to further European countries is planned through the Country Chapter model.

We regularly review our terminology and adapt it to the current state of scientific discourse.

Experience reports are collected through an anonymous, three-stage online form:

1. Context data: Industry (Audio, Film, Games), country (DACH, prospectively EU), experience type, and time period. These structured inputs enable subsequent filtering and pattern analysis.
2. Experience report: A free-text field with no minimum length, where contributors describe their experience in their own words. The length and level of detail are entirely at the discretion of the submitting person.
3. Consent and submission: Before submitting, explicit consent is obtained for publication of the anonymized contribution. A summary of the inputs allows a final review.

Moderation

Every submission is reviewed by an editorial review body before publication. Moderation comprises:

• Anonymization review: Removal or generalization of all information that could enable identification (company names, project titles, specific location details).
• Relevance review: Verification that the contribution falls within the thematic scope of the database (gender-specific experiences in the professional context of the creative industries).
• Quality assurance: Ensuring that the contribution complies with the project's ethical principles.

The technical processing of form submissions is handled through our own API on EU servers. Further information on data processing and data protection can be found in the privacy policy.

The database captures experiences in six scientifically grounded categories.

The database captures experiences from fourteen industry categories within the creative sector. Classification is self-reported by the submitting person. Overlaps are possible — when in doubt, choose the industry where the experience primarily took place.

The experience reports captured in the database are analyzed and visualized through an interactive dashboard. The analysis includes the detection of distribution patterns across experience types, regions, and industries.

It is important to note: the analysis is limited to descriptive statistics. No causal relationships are claimed and no statistical inferences are drawn. The dashboard shows distribution patterns of submitted experiences — not prevalence rates and not representative frequencies.

The visualizations serve pattern recognition: where do certain experience types concentrate? Which industries and regions are over- or underrepresented? These findings are exploratory and hypothesis-generating, not hypothesis-testing.

Data Source

The analysis draws on anonymized experience reports submitted through a structured survey instrument. Each record contains categorical variables for industry, region (ISO 3166-1 alpha-2), career level, and reporting period, as well as one or more discrimination types from a closed six-category system. Data are retrieved via a public API; all aggregations are computed directly from raw records.

Filtering

Two filter criteria are available: industry and region. Both are combined conjunctively (logical AND). A case is included in the filtered subset if and only if it satisfies both active filter conditions. All visualizations operate on the same filtered subset.

Counting Logic for Multi-Type Cases

A single case may document multiple forms of discrimination simultaneously. A uniform counting rule applies across all analyses: a case with $k$ types is counted once in each of the $k$ categories. Consequently, the sum of type-specific frequencies exceeds the total case count. Percentages refer to the total number of type mentions, not to the number of cases, and therefore cannot be normalized to 100% of cases. This design choice preserves the visibility of intersectional experiences rather than forcing single-category assignment.

Statistical Methods

The analysis is limited to descriptive statistics. No causal claims are made and no inferential conclusions are drawn.

For time-series analysis, a linear trend is estimated using Ordinary Least Squares (OLS):

$$\beta = \frac{\sum_{i=1}^{n}(x_i - \bar{x})(y_i - \bar{y})}{\sum_{i=1}^{n}(x_i - \bar{x})^2}, \qquad \alpha = \bar{y} - \beta\bar{x}$$

where $x_i$ denotes the reporting year and $y_i$ the corresponding case count. The coefficient of determination

$$R^2 = 1 - \frac{\sum(y_i - \hat{y}_i)^2}{\sum(y_i - \bar{y})^2}$$

is reported for each trend line. The current calendar year is included in the regression and may bias the trend estimate downward when few cases have been recorded; the $R^2$ value serves to assess goodness of fit. Records with non-specific time references (epoch ranges) are excluded from the time-series analysis.

Co-occurrence analysis maps pairwise frequencies as a symmetric matrix. For each type pair $(i, j)$, the count of cases documenting both types simultaneously is recorded. The diagonal contains the total frequency per type.

Visualization Conventions

In heatmaps, cells with values below a configurable threshold ($n < 3$) are optionally masked and excluded from color-scale computation. In the co-occurrence matrix, masking applies only to co-occurrence cells, not to the diagonal. The current calendar year is marked separately in time-series charts. Small-multiple panels share a common y-axis scale derived from the filtered data to ensure visual comparability. The co-occurrence matrix is displayed only when the filtered subset contains at least 20 cases.

Reproducibility

The complete source code for all computations is publicly available (open source). The API delivers exclusively unprocessed individual-case data. All aggregations and statistical calculations are therefore independently reproducible.

The following limitations apply in addition to the general methodological constraints and must be considered in any interpretation of the dashboard visualizations:

• Convenience Sample / Self-Selection: The data originates exclusively from individuals who actively submitted an experience report. Individuals without discrimination experience or those who do not wish to share their experiences are not represented. The data does not represent a total population.
• Multi-Type Inflation: Since cases can document multiple discrimination types simultaneously, the sum of type counts is higher than the case count.
• Small Samples: With few cases per category, individual submissions can significantly alter the visualizations.
• Temporal Incompleteness: The current calendar year naturally contains fewer cases than completed years.
• Epoch Ranges: Cases with non-specific time references do not appear in the time-based charts.
• Geographic Scope: The database currently captures only experiences from the DACH region (Germany, Austria, Switzerland).
• No Causal Claims: The dashboard shows distribution patterns, not cause-and-effect relationships.

This data does not replace representative salary surveys. It documents individual experiences.

The methodological limitations of this project must be transparently stated:

• Self-Reported Data: All experience reports are based on the subjective perception of the submitting individuals. No independent verification of the reported circumstances takes place.
• Selection Bias: Individuals who have experienced inequality are more likely to be motivated to submit a contribution. The database therefore does not represent the entirety of experiences in the creative industries.
• Small Sample Size: Particularly during the development phase, the case count is low. Generalizations to the entire industry are not permissible.
• No Claim of Representativity: The database makes no claim to statistical representativity. It is an instrument for making individual experiences visible, not a substitute for standardized surveys.

A significant methodological limitation of the current data collection is that each experience report represents a snapshot. Structural discrimination patterns, however, can only be fully understood when individual trajectories are documented over time.

Classical longitudinal studies require the identifiability of participants, which contradicts this platform's promise of anonymity. As a planned methodological extension, a privacy-compliant mechanism is therefore being developed.

Such a procedure would, for the first time, enable a quantitative analysis of progression patterns of professional discrimination in the European creative industries.

The research gap in systematic longitudinal documentation of gender-based professional disadvantage is considerable. Existing studies predominantly capture prevalence, not individual trajectories.

The project follows the ethical principles of the Ethics Code of the German Sociological Association (DGS). The following principles guide data collection, processing, and publication:

Trauma-Informed Design

The submission of experience reports on discrimination and harassment requires a particularly sensitive approach. The submission form integrates the following protective measures:

• Content warnings: Before beginning the submission, users are informed about the potentially distressing nature of the content. Crisis resources (helpline, women's helpline) are provided.
• Safe exit: A "Leave page" function enables immediate departure from the page.
• Own pace: No time pressure, no mandatory fields in the free-text area, no minimum length for experience reports.

Anonymization

Protecting the identity of contributing individuals has the highest priority:

• Direct identifiers (name, email, company name, exact dates) are never collected.
• Quasi-identifiers (e.g., year references, location details) are generalized before publication.
• Narrative review: Every free-text contribution is editorially reviewed for identifying details and adjusted if necessary.
• k-Anonymity: Filter combinations yielding fewer than five cases are not displayed individually, to prevent inferences about specific individuals.

Informed Consent

Before submitting, explicit consent is obtained. Contributors are informed about the purpose of data collection, publication as an anonymized contribution, and data processing.

Since no personal data is collected, individual submissions cannot be attributed to specific individuals after submission. Revocation or deletion of individual contributions is therefore not possible after publication. This circumstance is transparently communicated during the consent process.

Moderation

All submissions are reviewed by an editorial review body before publication. Moderation follows a standardized anonymization checklist and reviews each contribution for identifying details, thematic relevance, and compliance with ethical principles.

Moderator Protection

Reading experience reports about discrimination and harassment can trigger secondary traumatization. The project acknowledges this burden and provides protective measures for moderating individuals: time limits on moderation work, peer reflection, and clear protocols for particularly distressing content.

Detailed information on data protection, data processing, and the rights of affected individuals can be found in the privacy policy.

An interdisciplinary scientific advisory board from the fields of sociology, labor law, and gender studies is currently being assembled. If you are interested in participating, please contact info@moiree.eu.