Study Under Review:
Hough, K. L. (2014). Internal accountability and district achievement: How superintendents affect student learning. Journal of School Leadership, 24(1), 1-30. https://doi.org/10.1177/105268461402400102
Hough's 2014 study examines a question bridging leadership psychology and district governance: whether superintendent self-awareness, specifically, whether superintendents accurately perceive their own accountability behaviors relative to how subordinate administrators perceive those behaviors, predicts district student achievement. The answer, it turns out, is yes, and with effect sizes that demand attention. For a full explanation of the PEARLS framework used in this review, see our methodology page.
What the Study Found
Hough's 2014 study examines whether superintendent self-awareness, specifically, how accurately superintendents perceive their own accountability behaviors relative to how subordinate central office administrators perceive those same behaviors, predicts district student achievement. The answer is yes, and with effect sizes that are unusual in governance research.
Superintendents and their subordinate administrators completed surveys rating superintendent accountability behaviors. Hough then classified superintendents into three categories based on the gap between self-ratings and subordinate ratings: those who agreed with subordinates' assessments, those who overestimated their behaviors relative to subordinates, and those who underestimated them. After controlling for district socioeconomic status, the self-other agreement category alone explained 22% of the variance in district mathematics scores and 19% in reading, effect sizes that are rare in the school governance literature.
The direction of the effect matters. Superintendents who overestimated their accountability behaviors led districts with significantly lower achievement than those who agreed with or underestimated their subordinates' ratings. Underestimation did not carry the same negative association, suggesting the problem is not disagreement per se but a specific pattern of inflated self-perception.
Why the Self-Assessment Gap Matters
The self-other rating agreement framework Hough applies is not new. It is drawn from decades of organizational psychology research showing that overestimators tend to be less effective leaders across a range of industries and institutional contexts. What Hough's study does is extend that well-established pattern into K -- 12 district leadership and connect it to student achievement outcomes. That extension is consequential.
The study cannot resolve the causal direction, three explanations are consistent with the data. Superintendent self-awareness may cause better leadership behaviors, which cause better outcomes. High-performing district cultures may generate accurate feedback environments, producing better agreement between superintendent and subordinate ratings. Or some third variable, experience, board quality, district stability, may cause both accurate self-assessment and better outcomes. The cross-sectional design cannot distinguish among these. What it can establish is that the size of the association is strong enough that the self-other agreement construct deserves serious diagnostic attention regardless of which causal pathway turns out to be primary.
The specific behavioral domain matters too. The study measured accountability behaviors, setting expectations, monitoring progress, following through. These are the behaviors most directly connected to the governance-management interface. They are also the behaviors where self-delusion may be most costly, because a superintendent who believes accountability processes are more robust than they are has less incentive to strengthen them.
What Boards Should Take From This
Boards conducting superintendent evaluation typically rely on their own observations, superintendent self-reports, and student outcome data. This study suggests a diagnostic layer is missing: the perspectives of subordinate central office administrators on the superintendent's actual accountability behaviors. That gap, when it surfaces as systematic overestimation, appears to be a meaningful signal about district performance.
- Build structured multi-rater feedback into superintendent evaluation: Boards that rely exclusively on their own assessments and the superintendent's self-reports are missing the comparison that this study identifies as consequential. Central office administrator perspectives on superintendent accountability behaviors are worth collecting as a standard component of evaluation, not as a punitive instrument but as a calibration tool.
- Treat overestimation as a diagnostic flag, not a character indictment: The finding is that overestimation correlates with lower district performance. Boards should approach this signal as information, an indicator that calibration work may be needed, not as evidence of bad faith or incompetence. The response is coaching and development, not discipline.
- Focus evaluation on accountability behaviors specifically: Board evaluation processes that weight relational and communication dimensions heavily while underweighting accountability rigor, setting expectations, monitoring, following through, may be missing the behavioral domain this research identifies as most consequential. The construct the study measured maps directly onto what boards are trying to assess when they evaluate superintendent effectiveness.
- Weight the evidence appropriately: The effect sizes are large and the theoretical grounding is solid. But the cross-sectional design, undisclosed sample details, and reliance on self-report measures mean this study warrants careful practice change, not immediate structural overhaul of evaluation systems. It is a strong rationale for piloting multi-rater feedback while the longitudinal evidence base develops.
Hough's 2014 study identifies a striking association between superintendent self-awareness and district student outcomes, one with effect sizes large enough to take seriously even in the presence of design limitations. The finding that overestimating one's own accountability behaviors predicts lower district achievement gives boards a specific, actionable diagnostic signal. The study cannot prove causation, but it provides well-grounded empirical support for incorporating multi-rater feedback into superintendent evaluation as a standard governance practice.