Modeling and interpreting regressions with interactions

IF 1.1 Q3 BUSINESS, FINANCE

Journal of Accounting Literature Pub Date : 2019-06-01 DOI:10.1016/j.acclit.2018.08.001

Jeffrey J. Burks , David W. Randolph , Jim A. Seida

引用次数: 28

Abstract

This study examines the use of linear regressions that include interaction terms, finding frequent interpretation errors in published accounting research. We provide insights on how to estimate, interpret, and present interactive regression models, and explain seldom-used but easily-implemented methods to report conditional marginal effects. We also examine the use of interaction terms in tax and financial reporting trade-off studies, evaluating the conceptual fit between a regression model with interactions and alternative definitions of trade-off. Although we advocate the use of interactive models, noise levels common in accounting research greatly reduce the ability to detect interaction effects.

查看原文本刊更多论文

用交互建模和解释回归

本研究考察了线性回归的使用，包括相互作用的条款，发现频繁的解释错误发表的会计研究。我们提供了如何估计、解释和呈现交互回归模型的见解，并解释了很少使用但易于实施的方法来报告条件边际效应。我们还研究了在税收和财务报告权衡研究中相互作用术语的使用，评估了具有相互作用的回归模型和权衡的替代定义之间的概念契合度。虽然我们提倡使用互动模型，但会计研究中常见的噪音水平大大降低了检测互动效应的能力。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Journal of Accounting Literature BUSINESS, FINANCE-

CiteScore

6.50

自引率

0.00%

发文量

期刊介绍： The objective of the Journal is to publish papers that make a fundamental and substantial contribution to the understanding of accounting phenomena. To this end, the Journal intends to publish papers that (1) synthesize an area of research in a concise and rigorous manner to assist academics and others to gain knowledge and appreciation of diverse research areas or (2) present high quality, multi-method, original research on a broad range of topics relevant to accounting, auditing and taxation. Topical coverage is broad and inclusive covering virtually all aspects of accounting. Consistent with the historical mission of the Journal, it is expected that the lead article of each issue will be a synthesis article on an important research topic. Other manuscripts to be included in a given issue will be a mix of synthesis and original research papers. In addition to traditional research topics and methods, we actively solicit manuscripts of the including, but not limited to, the following: • meta-analyses • field studies • critiques of papers published in other journals • emerging developments in accounting theory • commentaries on current issues • innovative experimental research with strong grounding in cognitive, social or anthropological sciences • creative archival analyses using non-standard methodologies or data sources with strong grounding in various social sciences • book reviews • "idea" papers that don''t fit into other established categories.