Variable selection for classification and forecasting of the family firm's socioemotional wealth
Identifiers
Permanent link (URI): http://hdl.handle.net/10017/58908DOI: https://doi.org/10.1002/for.3008
ISSN: 0277-6693
Date
2023-06-30Funders
Fundación Cajamurcia
Bibliographic citation
Journal of Forecasting, 2023, v. 42, n. 8, p. 2063-2078
Keywords
Estimated classification error
Family firms
Financial database
Random forests
Socioemotional wealth
Document type
info:eu-repo/semantics/article
Version
info:eu-repo/semantics/publishedVersion
Rights
Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)
Access rights
info:eu-repo/semantics/openAccess
Abstract
Socioemotional wealth (SEW) refers to those family-centered goals that are likely to have a major influence on the strategic decision-making process and performance of family firms. Many studies have used indirect indicators related to family involvement in ownership and management to measure SEW; meanwhile, others have developed scales to directly measure the level and importance of SEW in family firms. Limitations of both indirect and direct measures of SEW lead empirical research on SEW to be under threat. In the current study, we use random forests to identify the important indicators related to financial and economic decisions, as well as family-related measures, for explaining the family firms' SEW and to design a good prediction model using the smallest set of nonredundant indicators. Our results show that the model that exhibits the minimum out-of-bag sample (OOB) error rate includes variables that refer to the presence of family members in the firm's management positions, long-term nonfinancial debt, personnel expenditures, longterm financial investments, short-term financial debt, average storage period, and accounts receivables. For prediction, the model with a reasonably low estimated classification error includes only three variables, which refer to the presence of family members in the firm's management positions, long-term nonfinancial debt, and accounts receivables.
Files in this item
Files | Size | Format |
|
---|---|---|---|
Variable_selection_JoF_2023.pdf | 534.0Kb |
|
Files | Size | Format |
|
---|---|---|---|
Variable_selection_JoF_2023.pdf | 534.0Kb |
|
Collections
- ECODIR - Artículos [94]