The best schools in the country?

This post was revised on August 13th to take into account errors found in the second figure. The errors do not have any consequences for the main conclusion.

Oslo SV has chosen «More teachers – fewer tests» as the slogan for our efforts to make a better school for Oslo. I’m not going to write about that today. Rather, I’ll begin my writings on this topic with a post that shows how the right-wing parties’ portrayal of the school in Oslo hits off the mark.

There is much good to be said about the public schools in Oslo, but that’s not to say that its virtues are the results of Høyre’s policies or that they’re very much better than schools elsewhere in Norway. Høyre, however, likes to boast that Oslo’s schools are the country’s best, even the world’s best  — no less. Can this be right?

This, obviously, is a very difficult question to answer. How do we ascertain that one school is better than another? Is the best school the one where students’ mean grades are the highest, or the one which instils the highest self esteem in its graduates? Do any of our measuring tools gauge how pupils «evaluate the consequences their actions have for others, and judge them with ethical consciousness», as the Norwegian Education Code stipulates?

I won’t try to tackle these problems here. Instead, I’ll perform two very simple analyses based on criteria that Høyre likes to use. It turns out that the results of two of the most important tests in Norwegian Elementary school hardly support a claim that Oslo’s schools are the best in Norway, let alone the World.

Oslo, it turns out, is not the average Norwegian county. One of the factors distinguishing Oslo from the other counties is that its labor market shows a high demand for workers with higher education.  I’m not going to discuss whether this is the reason why more people in Oslo have higher education, beyond pointing out this fact.

Keeping this in mind, and knowing also that the parents’ level of education plays a major role for their children’s academic achievements, let’s look at some numbers — mean scores on final exams in the Norwegian elementary school by county:

Final exam results for 10th grade students as a function of the average number of years of education in the parents' generation. Source: As described in the post's final paragraph.
Final exam results for 10th grade students as a function of the average number of years of education in the parents’ generation. Source: As described in the post’s final paragraph.

Where in this figure do we find Oslo? At the far right. Quite clearly, pupils at Oslo’s schools score better on their 10th grade exams than the national average. However, the figure also shows the huge gap between the parent generation’s highest completed education in Oslo and in the rest of Norway. If we make a model that corrects the exam scores for the number of years the parent generation went to school in average (the straight lines in the figure), Oslo doesn’t perform better than expected. On the contrary, the scores are exactly what we would expect them to be.

How about the national tests («nasjonale prøver» in Norwegian)? We’re frequently told that Oslo’s schools score particularly well here:

Update August 13th: Due to an error in the sorting of the data used for the original plots, the original post found no correlation between the parent generations’ level of education and National Test scores. The figure has been updated and verified.

National Test scores per county as a function of the mean number of years of education in the parent generations. Source: UDIR / SSB
National Test scores per county as a function of the mean number of years of education in the parent generations. Oslo’s scores in the upper right portion of the plot. Source: UDIR / SSB

However, as the graph shows, the county-wide mean for Oslo does not fall outside the bounds of statistical prediction when the parent generations’ level of education is taken into account. The plots highlight the fact that the National Tests are not able to uncover any difference in educational achievement that is not explained by the parents’ level of education.

Some final words: There are many good things to say about Oslo’s schools. They are the workplace of highly qualified and dedicated teachers. Acknowledging this, however, shouldn’t stop us from questioning how Høyre is marketing its alleged results in its showcase county. The analyses presented here are unpretentious, but they highlight some important issues that are rarely raised when Oslo’s schools are discussed.

The statistics on educational levels (table 08921, numbers from 2014) and population numbers (table 03031, numbers from 2014) were collected from Statistics Norway.  Final exam scores and national test scores were collected from the The Norwegian Directorate for Education and Training (preliminary final exam scores (2015), national test scores (2014)).