Types of Homeschool Environments and Need Support for Children’s Achievement Motivation

ABSTRACT

Working within a self-determination theory (SDT) framework, this study used cluster analysis to examine the naturally occurring types of homeschool-learning environments parents (N = 457) have created. Measures of support for student autonomy, mastery goal structure, and use of conditional regard were adapted for a homeschool context and used as constituting variables. Follow-up measures of parental need satisfaction, efficacy, student academic engagement, teaching practices, and demographics were used to identify significant differences among homeschooling motivational profiles. A five-cluster solution best fit the data: a high need support profile, low need support profile and three profiles of mixed need support. In general, the high need and mixed need support profiles were associated with higher student engagement, need satisfaction, efficacy for homeschooling, and frequent use of teaching strategies that promote autonomous motivation and support for student competence. The low need support profile was significantly associated with lower need satisfaction and teaching strategies associated with control. Higher levels of academic engagement were reported for those students homeschooled longer and at higher grade levels. Male teaching parents (N = 29) reported significantly less need satisfaction and were significantly more represented in the low need support profile. These findings point to the utility of self-determination theory for characterizing the motivational environments of homeschools.

KEYWORDS:

• achievement motivation
• homeschooling
• homeschool teaching practices
• self-determination theory

View correction statement:

Corrigendum

Notes

1. The slight control subscale of the PIS had a less than satisfactory reliability (a = .67). However, following Reeve’s recommended scoring procedure (Reeve, Bolt, & Cai, 1999), this subscale on the PIS is effectively canceled out.

2. These correlations among the variables used for the cluster analysis also met the recommended relationship among variables for performing a MANOVA (e.g., high, negative correlations or moderate correlations in either direction; Tabachnick & Fidell, 2007).