Despite its long history, extant research on the relations among budgetary participation, reliance on accounting performance measures (RAPM) and individual-level consequent variables remains marked by conflicting findings. To solve these conflicts, prior research has introduced variables that might mediate or moderate the relations. But many studies use small samples, and their conflicting findings might be due to statistical artefacts, such as sampling error. This paper undertakes a meta-analysis to assess whether prior findings are homogeneous after correcting for these artefacts and to explore the moderating effects of study design choices regarding construct measurement, random vs. non-random sampling and industry differences. Many relations of participative budgeting and RAPM are homogeneous; several are truly heterogeneous. Construct measurement emerges as the most important moderator: different measures of managerial performance explain the relation of participative budgeting to managerial performance, and different measures of RAPM largely clarify the relation of RAPM to budgetary participation. This relation also seems contingent on industry differences. These findings have important implications for research and managerial practice.
© 2001-2024 Fundación Dialnet · Todos los derechos reservados