Political Engagement Profiles in a Student Population: A Validity Study
Faculty Advisor Name
Dena Pastor
Department
Department of Graduate Psychology
Description
In 2012, the institution of higher education was called on by the United States Department of Education to strengthen their emphasis on students’ civic learning and democratic participation (National Task Force on Civic Learning and Democratic Engagement, 2012). In 2017, the State Council for Higher Education in Virginia introduced a policy that required higher education institutions in the state of Virginia to assess civic engagement. With the calls from the US Department of Education and SCHEV, as well as JMU calling itself the “engaged university,” it is imperative that we assess students’ political knowledge and engagement. After forming the James Madison Center for Civic Engagement (JMCCE), the Center for Assessment and Research Studies (CARS) was invited to collect data to help assess JMU’s and JMCCE’s efforts in getting students to become more politically engaged citizens.
Using the Political Engagement Project Survey (PEPS) (Beaumont et al., 2006), 708 JMU students were asked about various information surrounding their political identity. Two of the subscales of the PEPS given to students asked them their likelihood of engaging in about 15 different political activities in the future. Analyzing the patterns of responses using Latent Class Analysis, the presenter found four different types of students on JMU’s campus: high-engagement, voting-only, conventional-action, and nonconventional-action. The high-engagement group has a high probability (above .7) of participating in 13 of the 14 activities students were asked about. The disengaged group is noted as having a high probability of voting but has a low probability (below .3) of participating in almost every other action. The conventional-action class is characterized as having a higher probability of participating in activities like contacting a public official and working for a campaign, and having a lower probability of protesting and performing political consumerism Finally, the nonconventional-action class is defined by having high probabilities of protesting and performing political consumerism but has low probability of participating in contacting public officials and working with campaigns. Each of the four classes have defining features, but it is not enough to observe these groups at face-level. Validity evidence needs to be gained in order to give more meaning to these classes so they can be used appropriately in JMU’s efforts to create more engaged students.
Validity evidence was gained through finding literature that informed relationships between the four classes and auxiliary variables (variables collected but not used in the initial analyses). After deciding on four auxiliary variables (gender, race, political knowledge, and political affiliation), multiple hypotheses were formed to test the relationships between each external variable and how it related to the four classes. To research these hypotheses, multiple Chi-Square tests were performed. The focus of the presentation will be on the results of the validity study.
Political Engagement Profiles in a Student Population: A Validity Study
In 2012, the institution of higher education was called on by the United States Department of Education to strengthen their emphasis on students’ civic learning and democratic participation (National Task Force on Civic Learning and Democratic Engagement, 2012). In 2017, the State Council for Higher Education in Virginia introduced a policy that required higher education institutions in the state of Virginia to assess civic engagement. With the calls from the US Department of Education and SCHEV, as well as JMU calling itself the “engaged university,” it is imperative that we assess students’ political knowledge and engagement. After forming the James Madison Center for Civic Engagement (JMCCE), the Center for Assessment and Research Studies (CARS) was invited to collect data to help assess JMU’s and JMCCE’s efforts in getting students to become more politically engaged citizens.
Using the Political Engagement Project Survey (PEPS) (Beaumont et al., 2006), 708 JMU students were asked about various information surrounding their political identity. Two of the subscales of the PEPS given to students asked them their likelihood of engaging in about 15 different political activities in the future. Analyzing the patterns of responses using Latent Class Analysis, the presenter found four different types of students on JMU’s campus: high-engagement, voting-only, conventional-action, and nonconventional-action. The high-engagement group has a high probability (above .7) of participating in 13 of the 14 activities students were asked about. The disengaged group is noted as having a high probability of voting but has a low probability (below .3) of participating in almost every other action. The conventional-action class is characterized as having a higher probability of participating in activities like contacting a public official and working for a campaign, and having a lower probability of protesting and performing political consumerism Finally, the nonconventional-action class is defined by having high probabilities of protesting and performing political consumerism but has low probability of participating in contacting public officials and working with campaigns. Each of the four classes have defining features, but it is not enough to observe these groups at face-level. Validity evidence needs to be gained in order to give more meaning to these classes so they can be used appropriately in JMU’s efforts to create more engaged students.
Validity evidence was gained through finding literature that informed relationships between the four classes and auxiliary variables (variables collected but not used in the initial analyses). After deciding on four auxiliary variables (gender, race, political knowledge, and political affiliation), multiple hypotheses were formed to test the relationships between each external variable and how it related to the four classes. To research these hypotheses, multiple Chi-Square tests were performed. The focus of the presentation will be on the results of the validity study.
