Peer-mentoring of students in rural and low ses schools

9Results

Data were analysed using mixed-model analyses of variance (ANOVA), with effect sizes reported in Cohen's d (Cohen 1988). We were able to get four points of data for 46 students, 24 from the low-SES school and 22 from the semi-rural school. A median split was performed on the sample to determine those students who had received more intensive amounts of mentoring and those who had received low levels of mentor contact.

Likelihood of attending university and TAFE

Participants who had consistently higher levels of mentoring² had a higher score on intentions to go to university across all four time points than those who had received minimal or no mentoring.³ The difference in intention to attend university between students who had received consistently high levels of mentoring and those who had received minimal or no mentoring was similar at each time point, resulting in a non-significant time by mentoring interaction (p = .231). Figure 2 shows that students who received mentoring were consistently more inclined to attend university following high school graduation than those who received no such mentoring. The mentoring had no effect on intention to go to TAFE.⁴ Figure 3 shows the intention to go to TAFE over time.

Figure Students’ estimated likelihood they would attend university over time split by high and low levels of mentoring

Note: Error bars denote standard error.
Figure Students’ estimated likelihood they would attend TAFE over time split by high and low levels of mentoring

Note: Error bars denote standard error.

At each single time point, we had more students than in the complete analyses due to drop-outs and late enrolments. At each time point we conducted an analysis of how likely participants were to attend university, split by how much mentoring they had received. Participants at time 1 indicated that they were more likely to attend university when they had received higher levels of peer mentoring.⁵ At all other points in time, participants were not significantly more likely to attend university when mentored, although marginal trends were observed at time 2⁶ and time 4.⁷ This may indicate that, for mentoring to be successful, it must be ongoing, as the primary analysis reveals a significantly higher level of intention to go to university among those students who received mentoring across all four time points.

University in-group identification: students who received high levels of mentoring identified themselves as significantly more affiliated with university students as an in-group⁸ than those who received low levels of mentoring.⁹

Financial burden: there was a significant main effect of time on predictions of the financial burden of attending university, in which, over time, all students tended to estimate the costs of attending university as lower, irrespective of mentoring.¹⁰ This represented a small to moderate decline in cost estimates from time 1 to time 4.¹¹ As there was no effect of mentoring on cost perceptions, this indicates that the effect of mentoring was not produced through the mentors affecting the students’ perceptions of cost. As cost perceptions tended to regress to the mid-point across the four time points, it may be that students’ perceptions tend to become more closely aligned to the actual cost of university as they increase in age and gain an appreciation of the costs of goods and services.

Understanding of university: there was no significant effect of mentoring or time on self-reported understanding of university and what occurs at a university.

Cognitive distance: there was a marginal tendency for students who had received high levels of mentoring to believe that Flinders University was physically closer to them¹² than students who had received little or no mentoring.¹³

Mediation analyses: a multiple linear regression showed that, when controlling for the effect of which school participants attended, mentoring had a significant effect on cognitive distance.¹⁴ After controlling for in-group identification and intent to attend university, a multiple linear regression showed that the effect of the mentoring on the perceived likelihood of attending university became non-significant.¹⁵ This indicates that the effects of the mentoring were completely accounted for by the reduction of cognitive distance and the increase of university in-group identification, as shown in figure 4. The total variance explained by the combined effect of cognitive distance and in-group identification, after controlling for the school participants attended, was 54.6% of the difference between the mentored and non-mentored group. It is highly likely that a predisposition towards attending university and other factors account for the remainder of the variance.

Figure The effect of mentoring through cognitive distance and in-group identification

In late November 2011, a further step, in addition to collecting data via questionnaires, was conducted: five Year 10 students (one female) were interviewed as a group about their experience of being in the mentoring program. The interviews showed that visits from mentors and the interactions with them, mostly during school time but occasionally outside school (for example, at the local supermarket), did influence students’ thinking about what they intended to do when they left school. For example, ‘both mentors were doing teaching [and] it seemed like fun … this helped me because this is what I want to do’. Another student found the mentor’s advice and experience helpful, but in a way that indicated to them what they did not want to study: ‘[it’s] hard to work with little kids, too hard, not what I wanted to do … helped me realise what I didn’t want to do’.

In addition to specific career-oriented advice, the opportunities for visiting the Flinders University campus were positively endorsed by each student, encapsulated in comments such as ‘a good introduction to university, had no idea what it was like before going’; ‘opens your mind to what you can do’; ‘can see you can do what you like — yeah’. The group was also clear about how they would like to see the mentoring program develop: ‘more visits to schools by mentors’; organise it ‘so that everyone [in the school] gets the experience’; ‘have some mock lectures … so you know what is coming’; and more hands-on activity during visits to the university.