Freedom of information legislation as a tool for obtaining meaningful law school grade information

 

 “Standard deviation” is a commonly used measure of statistical dispersion. Statistical dispersion refers to how widely spread the values in a data set are. If the data points are close to the mean (the average) the standard deviation is small (e.g. class B). Conversely, if many data points are far from the mean then the standard deviation is large (e.g. class A). To know how good a particular grade is, some indication of dispersion is required.

 

For some classes (depending on class size and other factors) UBC law “shifts” the grades of the entire class to ensure that the class average is within a specified range. However, law school grades are often not normally distributed (i.e. they do not have a neat bell shaped curve). I understand that UBC does not distort the distribution of grades to make it normally distributed. Therefore, calculating standard deviations is not a meaningful way of determining the dispersion of grades.

 

However, individual class ranks are a useful indicator of student performance because they provide an indication of the dispersion of grades from the class average. I was personally told by two of my law school professors that they “do not give out high grades”. In these courses one would expect a narrow distribution of grades and so individual class rankings are particularly important for assessing performance: a grade a few percentage points above class average may actually be near the top of the class.  

 

I applied for individual class rank information for the classes I took at UBC, and eventually got it. Tables 1 and 2 below confirm the “variable dispersion” effect described above. Comparing my marks for Trusts (80%) and Charter (75%), one would think that I did better in Trusts than in Charter. In fact, I did worse in Trusts: 19.1% of the class beat me in Trusts, but only 17.6% of the class beat me in Charter. A comparison between Intellectual Property and Secured Transactions in third year illustrates the same effect.

 

 

These examples illustrate that one cannot meaningfully compare grade information between classes unless those classes have similar dispersions. Since law school classes do not naturally have similar dispersions, and no fitting to a curve is done (as distinct from shifting the curve, whatever its shape, to achieve a particular class average), providing individual class ranks is the most practical way of providing meaningful grade information.

 

The Freedom of Information and Protection of Privacy Act, R.S.B.C. 1996, c. 165 [the Act] provides the public in British Columbia with a broad right to disclosure of records held by public bodies.

 

Section 3 of the Act says that, generally, it applies to all records in the custody or under the control of a public body. Section 4 of the Act provides individuals with the right to obtain “any record”:  

 

4(1) A person who makes a request under section 5 has a right of access to any record in the custody or under the control of a public body, including a record containing personal information about the applicant.

 

Division 2 of Part 2 (ss. 12 to 22.1) of the Act lists restrictions on the scope of disclosure, and prevents a public body from disclosing private information of other persons. Therefore, producing records under s. 4 may require the public body to sever information from an existing record before disclosing it to the applicant.

 

As a general rule, public bodies are only required to disclose records already in their possession, they are not required to create records which the applicant would like, but which do not already exist. However, s. 6(2) of the Act does impose a limited duty on public bodies to create records which do not yet exist, so long as creating the requested record will not impose an onerous burden on the public body.

 

For a detailed discussion of the statutory provisions and case law dealing with the duty on public bodies to produce records under s. 4(2) and create records under s. 6(2), see http://www.legaltree.ca/node/715.

 

On June 2, 2006 I applied to UBC for disclosure of:

o       Overall class ranks for each year of law school.

o       Individual class ranks for each class I took at law school.

 

In an August 21, 2006 letter, UBC stated that it could provide my overall class rank for first year, but not for later years because they said they had not calculated overall class ranks since the 2003/2004 academic year.

 

UBC said that it had never calculated individual class ranks and would not provide that information.

 

After much further correspondence over many months, and the urging of the Office of the Privacy and Information Commissioner to try and resolve the matter, in a December 12, 2006 letter UBC provided overall class ranks for my second and third years. UBC maintained its position that it had never calculated individual class ranks and that it was not obliged to provide such information under the Act.

 

My position was that if it did have such information available, UBC was required by s. 6(2) of the Act to produce it. Alternatively, I said that UBC had complete class lists and was at the very least, under s. 4(2) of the Act, required to provide me with copies of those class lists after redacting the names of the other students, but leaving the grades of the other students showing. That way I could count off the number of students who had gotten higher grades than I had.

 

I referred the matter back to the Office of the Privacy and Information Commissioner. After communicating with UBC, a representative of the Commissioner said that to avoid the matter going to a hearing, UBC would provide me with redacted class lists which would allow me to count off my ranking i.e. they would provide redacted records under s. 4(2) of the Act, but would not create records under s. 6(2) of the Act. Because it would allow me to obtain the information I sought, I agreed to that offer.

 

In an October 11, 2007 letter UBC provided me with redacted class records showing the grades of all students in the class, but with the names of all students except me redacted.  This allowed me to read off my individual class rank information, including the data set out in Tables 1 and 2 above.  

  

The program UBC uses to manage student grade information is called the Student Information System (“SIS”). The SIS page of UBC’s website (http://www.students.ubc.ca/facultystaff/sis.cfm) provides information about the capabilities of the system. The SIS appears to be a sophisticated program and has a number of built in features for manipulating data. Furthermore, data can be exported from the SIS into Microsoft Excel where it can be further manipulated.

 

It its correspondence with me, UBC never said that the SIS cannot produce individual class rankings. It only said that the Faculty of Law “has never ranked its students on a class-by-class basis” (August 21, 2006; December 12, 2006; and April 27, 2007, letters) and that “Enrolment Services does not collect ranking information on individual classes through the SIS” (April 27, 2007, letter).

 

It may be that the SIS does not have the capability to create individual class ranks. However, once the data is in Excel, it can be manipulated using visual basic macros. As part of my submissions under s. 6(2) of the Act, I provided UBC with a visual basic macro which would calculate individual class rank using the class data exported from the SIS into Excel. 

 

Because the matter was settled it was never determined whether UBC has a duty to produce individual class ranks under s. 6(2) of the Act. Nor do I know whether the SIS had the ability to generate individual class ranks. However, it is arguable that UBC has a duty to create and produce such records: http://www.legaltree.ca/node/715.

 

Many students don’t care about having meaningful grade information. Some don’t even really want to be at law school. Others may prefer to have no grade information, but simply get a P for pass or F for fail, and rely on considerations other than grades when applying for jobs. Such a system is used by some medical schools which likely leads to medical internships being allocated according to who did the most volunteer work, who got the best reference letters from their professors, and probably who got the best undergraduate grades.

 

I acknowledge that law school grades and exams are not an exact science, and that the correlation between grades and employee quality is not perfect. However, so long as employers are using grades to evaluate job applicants, students who spend thousands of dollars to attend law school should be provided with meaningful grade information.

 

In my view, to be meaningful, grade information has to account for the variable dispersions in the grades of individual classes. Calculating individual class ranks allows comparison of students in different classes independently of the distributions of grades or the class averages of the individual classes.

 

Regarding overall class rankings, because they are based on grades obtained in individual classes they are influenced by differences in class averages and grade distributions. This is problematic when comparing different students who are taking different sets of classes. A better indication of overall performance would be to use the individual class ranks to calculate an overall credit weighted percentile ranking. This would allow fair comparison of students regardless of which, or how many, classes they took. Again, many will not like the idea of reducing students to a single number. However, when considering candidates employers typically add up the applicant’s grades and divide by the number of classes taken to get an overall average percentage: also a single number, but one that is less reliable than a credit weighted percentile ranking.

 

Knowing that my application for individual class rank information succeeded, other students can apply for their individual class ranks. Perhaps if enough students apply for such information universities will make such information available to students who want it as a matter of course.