Writing questions about sensitive character themes

The substantive question

The published HEXACO-60 illustrates a deliberate effort to represent broad domain content in a shorter form. MSDC also contains sixty statements, but the similarity in length has no psychometric significance. Its questions are original and are organized into ten statements for each proposed domain, with two oppositely keyed statements in each group.

Character language creates special design problems. A question can invite idealized self-presentation if it reads like a public declaration of virtue. It can also imply accusation if it assumes wrongdoing. MSDC uses ordinary exchanges and manageable conflicts rather than dramatic moral dilemmas. Even then, the questions need interviews to establish how respondents understand an unjustified benefit, a reasonable explanation, or a minor offence. These phrases require judgment and may differ across settings. Emotional sensitivity items should not imply that seeking support is weakness. Patience items should not encourage tolerating mistreatment. A content review must inspect these implied messages as carefully as the scoring direction.

A worked interpretation

A statement about returning an accidental benefit may evoke a trivial overpayment for one reader and an ambiguous reward policy for another. Their answers could differ because the imagined situation differs. A pilot interview can ask for the example without requiring disclosure of sensitive details.

Similarly, letting an irritation settle is different from failing to enforce a boundary. If participants regularly confuse those meanings, the item should be rewritten or removed. The research team should document the reason, then reevaluate content coverage. Keeping an item because it increases a coefficient would not justify retaining a harmful or misleading interpretation in the report.

A question worth testing

Questions about advantage should include enough context to make the relevant judgment possible without demanding confession. If participants disagree about whether a benefit was unjustified, their responses may reveal different assumptions rather than different conduct. A development interview can investigate the assumptions with hypothetical examples. The team should revise ambiguous wording before considering statistical item removal, because the content problem may be more important than the observed correlation.

How we would evaluate MSDC

MSDC's question review should distinguish moral salience from useful measurement content. A dramatic statement about dishonesty might produce easy agreement with the desirable response while offering little insight into ordinary choices. An everyday statement about acknowledging an accidental benefit may be more discussable, but it still needs clear assumptions. Reviewers would ask whether participants can answer from normal experience without revealing confidential events or feeling accused. The form should not solicit admissions merely to make its character language appear serious.

Reverse wording deserves particular scrutiny in this setting. Strong disagreement with an unfavorable statement can reflect comprehension, self-presentation, a genuinely different tendency, or lack of relevant experience. MSDC cannot distinguish those possibilities automatically, and it should never label the respondent deceptive on that basis. Interviewers would explore how respondents interpreted the benefit, disagreement, or reassurance described, using hypothetical examples when privacy matters. The item map should balance situations across each domain and preserve the difference between fair dealing and compliance with authority. A proposed question that rewards obedience regardless of the rule's legitimacy would not fit the intended construct simply because it correlates with the remaining items. Content judgment must constrain statistical selection.