How are upper secondary school students’ expectancy-value profiles associated with achievement and university STEM major? A cross-domain comparison

• Published in: Contemporary educational psychology 2019-07, Vol.58, p.149-162
• Main Authors: Gaspard, Hanna, Wille, Eike, Wormington, Stephanie V., Hulleman, Chris S.
• Format: Article
• Language: English
• Subjects: Achievement; Expectancy-value theory; Latent profile analysis; University major
• ISSN: 0361-476X; 1090-2384
• DOI: 10.1016/j.cedpsych.2019.02.005

•Expectancy-value profiles across math and English were examined.•Latent profile analyses revealed four distinct profiles.•Students’ demographic characteristics were meaningfully related with these profiles.•Students in the four profiles differed in their math and English achievement.•Profile membership predicted later STEM major choice. Expectancy-value theory (Eccles, 2009) posits that students’ relative expectancies and values across domains inform their academic choices. Students should therefore be more likely to choose a STEM major if they have higher expectancies and values in STEM domains compared with other domains. Accordingly, this study aimed to explore how upper secondary school students’ profiles in expectancy-value beliefs in math and English are related to concurrent achievement and university major choice. Data on expectancies and values in math and English were collected from 2153 German students in their last school year, along with their concurrent math and English achievement and their university major 2 years later. Latent profile analyses revealed four distinct expectancy-value profiles characterized as Low Math/High English, Moderate Math/Moderate English, High Math/Low English, and High Math/High English. Students’ gender, socioeconomic status, and type of school were meaningfully associated with profile membership. For instance, female students were overrepresented in the Low Math/High English profile compared with other profiles. Students in the four profiles also differed in their math and English achievement. These differences were mostly in line with students’ expectancies and values in the respective domain, but some differences suggested that intraindividual cross-domain comparison processes were also at play. Finally, profile membership predicted students’ choice of a STEM major over and above demographic characteristics and achievement. Students in the High Math/Low English profile were most likely to choose a STEM major. These findings support the importance of considering intraindividual comparisons of expectancies and values for students’ achievement-related behavior and choices.