GM503 Unit 9 Translation Agency Trust Analysis Example

Errors Reported to One Manager Only: A Unit 9 Ethics and Trust Analysis at a Translation Agency

[Student Name]

Purdue University Global

GM503: Leadership Theory and Practice II

Unit 9 Ethics and Trust Analysis

[Instructor Name]

[Date]

The agency, its linguists, the vendor manager and every figure are composite and illustrative; the frameworks and studies cited are real.

The Agency and the Shift to Post-Editing

The agency translates technical manuals, software interfaces and medical device labeling for about sixty corporate clients. It relies on roughly two hundred freelance linguists, each coordinated by one of five vendor managers who assign work, negotiate rates and handle quality issues. Three years ago the agency began sending most work through machine translation first, with linguists paid to post-edit the output rather than translate from scratch. Many linguists distrusted the change. Machine output can contain errors that read fluently, such as a reversed negative in a dosage instruction, and finding them requires as much care as translating.

The agency asked linguists to report serious machine-translation errors through a form, so the engineering team could improve the system. Over two years, the form received 412 reports. Of these, 318 came from linguists coordinated by one vendor manager, Ingrid Sorensen, who works with about forty of the two hundred linguists. The other four managers' linguists submitted 94 reports among them.

Trust as Willingness to Be Vulnerable

Roger Mayer, James Davis and David Schoorman treated trust as a choice to accept exposure: a person who trusts lets someone else's actions affect her, expecting that the other will behave well even though she cannot watch or control what they do. Three perceived qualities make that choice reasonable. Ability is competence in the matter at hand. Benevolence is the sense that the other party wishes the trustor well. Integrity is the sense that the other party keeps to principles the trustor accepts (Mayer et al., 1995).

Reporting a machine-translation error is an act of vulnerability for a freelancer. It may suggest the linguist took too long on a job paid by the word, it may draw attention to a client's content and it depends on the manager's handling of the report.

Why Linguists Reported to Sorensen

Interviews with twelve linguists, six who worked with Sorensen and six who worked with other managers, point to all three components, but unevenly.

Ability. Sorensen is a former medical translator who reads three of the agency's main languages. Linguists said she understood why an error mattered. One described a report about a mistranslated contraindication that Sorensen escalated to the client within a day, explaining the clinical risk in terms the client's regulatory team accepted.

Benevolence. Linguists described her as protecting them. When a report revealed that a linguist had missed an error in an earlier job, Sorensen treated it as a system issue rather than a performance problem. Linguists working with other managers said they worried that reports would be used to cut their rates or reduce their assignments.

Integrity. Linguists said Sorensen told them what happened to each report, including when engineering declined to act, and that she had never promised something she did not deliver.

Linguists took the risk of reporting errors to the one manager they believed would use the reports to fix the system rather than to judge them.

The comparison suggests ability alone does not explain the difference: two of the other managers are also former translators. Benevolence and integrity, as linguists perceived them, distinguished Sorensen.

Ethical Leadership

Michael Brown, Linda Treviño and David Harrison described ethical leaders as people who behave appropriately in their own conduct and relationships and who encourage the same conduct in others by talking it through both ways, rewarding it and building it into decisions. Their studies linked that kind of leadership to followers' readiness to tell a supervisor about trouble (Brown et al., 2005). Sorensen's practice of explaining what happened to each report is a clear instance of two-way communication about a matter with ethical weight, since some errors could harm patients.

The Rate Cut

Last year the agency's executives announced a 12 percent cut in post-editing rates, effective in sixty days. Vendor managers had no say. Sorensen argued against the cut in two meetings and lost.

What she did next matters for the trust analysis. She called each of her forty linguists before the announcement, told them the cut was coming, said she had opposed it and explained what she could and could not do: she could not change the rate, but she could prioritize them for work in language pairs paid at a separate, uncut rate.

What Happened to the Reports

In the three months after the cut, reports from other managers' linguists fell to almost nothing, from about eleven a quarter to two. Reports from Sorensen's linguists fell too, from about forty a quarter to twenty-six, then recovered to thirty-four in the following quarter.

The interviews suggest why. Linguists who worked with Sorensen said the cut damaged their trust in the agency but not in her. Her early call and her honesty about her limits preserved their sense of her integrity, and her efforts to steer better-paid work to them preserved their sense of her benevolence. The dip reflected their anger at the agency; the recovery reflected their continued trust in the person receiving the reports.

A meta-analysis of trust in leadership found that trust in a direct leader relates more strongly to job-level outcomes, such as performance and helping behavior, than trust in organizational leadership does (Dirks & Ferrin, 2002). The agency's experience fits: linguists kept reporting errors to a direct manager they trusted even as their trust in the agency fell.

Implications

The agency has been treating error reporting as a technical process. The analysis suggests it is a trust process. Other vendor managers could adopt Sorensen's practices: telling linguists what happened to each report, separating error reports from performance reviews in writing and explaining decisions such as rate changes before they take effect. The agency could also measure what it now ignores: the share of each manager's linguists who submit at least one report a year, which would show where trust is weakening long before errors reach a client.

Conclusion

Linguists reported machine-translation errors mainly to one vendor manager because they trusted her benevolence and integrity, not only her ability. After a rate cut she opposed and could not stop, reports to her dipped briefly and then recovered, while reports to other managers nearly stopped. Mayer, Davis and Schoorman's framework explains the difference, and the case shows that trust in a direct manager can survive a loss of trust in the organization above her.

References

Brown, M. E., Treviño, L. K., & Harrison, D. A. (2005). Ethical leadership: A social learning perspective for construct development and testing. Organizational Behavior and Human Decision Processes, 97(2), 117-134. https://doi.org/10.1016/j.obhdp.2005.03.002

Dirks, K. T., & Ferrin, D. L. (2002). Trust in leadership: Meta-analytic findings and implications for research and practice. Journal of Applied Psychology, 87(4), 611-628. https://doi.org/10.1037/0021-9010.87.4.611

Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational trust. Academy of Management Review, 20(3), 709-734. https://doi.org/10.2307/258792