Students’ own low expectations can reinforce their disadvantage
Your expectations are likely to be formed in part by aspirations or what you would like to occur and the constraints that might get in the way. Constraints may be internal or external and all are assessed with a fair degree of uncertainty, which increases as the time horizon expands. Tying expectations and aspirations together is the concept of self-efficacyor “people’s beliefs in their ability to influence events that affect their lives”.
Studying expectations in addition to actual outcomes can provide a useful lens through which we can view class and intergenerational inequality. Low expectations can signal an early disengagement from important domains. They can be self-fulfilling. Why put in any effort if you don’t expect to benefit eventually from that effort?
Finally, if perceived constraints are not distributed evenly across the population, there will remain groups that don’t achieve what they might otherwise be able to or want to.
Along with Timothy Cameron, I have written on the role of expectations in explaining the education outcomes of Indigenous youth in Australia. Joanna Sikora and Larry Saha have considered these issues from both a gender and international perspective. This and other research shows that many groups in Australia have relatively low expectations.
What causes these differences, however, is a little more difficult to untangle. Some data from the Longitudinal Surveys of Australian Youthor LSAY will highlight this complexity.
Let’s begin by looking at the distribution of education expectations by three key dimensions: sex, Indigenous status and school sector. I categorise around 14,000 students into five groups based on expectations of high school completion and post-school study.
The majority of students expect eventually to complete a degree. This might be over-optimistic, but what is important is that these expectations vary significantly between groups.
48% of males in the sample expect to complete a degree compared to 63% of females. Indigenous students are much less likely to expect to complete a degree compared to the rest of the sample, with just 31.5%. Within school sectors, 48% of those in government schools expect to complete a degree compared to 65% of Catholic school students and 73% of independent school students.
Many potential reasons exist for this variation in expectations. For some, low expectations may fit an objective assessment of their own academic achievement to date and future academic potential. There is some truth to this in the data with a strong correlation between expectations and standardised test scores.
But what matters from a policy point of view is whether some groups’ expectations are too low given their observed academic ability. That’s not to say that these groups should be ignored if it turns out low expectations on average are driven by low measured academic achievement. But, in this case, it is the academic achievement that should be the focus of policy – for example, through early childhood education or school experience.
One way to test for this is through statistical techniques that hold other observable characteristics constant. The next figure looks at the difference in the probability of expecting to complete a degree for a few comparison groups:
- Those who live in a provincial or a remote area (compared to those in a major city);
- Females (compared to males);
- Indigenous students (compared to non-Indigenous students);
- Those who attend a Catholic school or an independent school (compared to a government school).
The final bar in the figure gives the estimated difference in the probability from a person’s parents having an extra year of schooling. Importantly, these differences are what we observe after controlling for a student’s age, English language background, socioeconomic status, school resources and test scores. We are comparing like with like.
Results summarised in the above figure fit well with other quantitative and qualitative research. Even with the same level of measured academic ability, those outside our major cities, and in particular Indigenous youth, are less likely to expect to complete a degree. Females and those outside the government school system are more likely to expect to do so.
Finally, parental education appears to have a large association. This is so even when you control for the potential direct effect of parental education on academic achievement.
If these differences in expectations translate into differences in outcomes (and the evidence suggest they do), then this opens up a potential avenue for the policy community to intervene. But, as Pip found out in Dickens’ masterpiece Great Expectations, managing expectations isn’t easy.
It is entirely plausible (though we don’t know this for sure) that some of those with low expectations are overly pessimistic because of their ethnicity, sex, location or school setting. Research suggests certain interventions can mitigate the salience of these background characteristics. Julio Garcia and Geoffrey Cohen summarise one promising approach:
Students are taught to attribute adversity and hardship to factors not directly relevant to race … Instead they are encouraged to attribute adversity and hardship to challenges inherent in school.
Such approaches fit within a broad category of interventions that Timothy Wilson has labelled the story-editing approach. These are described as: