On this page, information is given about the variables stored in the SweSAT database and the type of data they contain, how the coding has changed over the years, as well as time periods during which different variables have been in use.
This text is an overview of the information available in the SweSAT database. It is intended to serve as a support for researchers interested in requesting data from the database, and instead of naming specific variable names here, we focus on the information contained in the variables. When requesting data, the desired information and time period should be specified in our web form.
Data has been divided into the following three categories based on the information they contain:
Information related to the test administration includes, for example, information about the year and occasion (spring or autumn, indicated as 1986A and 1986B, for example) which has been reported in the same way since the test began, with a few exceptions. Information about where the test participants took their tests was indicated as counties until 1983 and then as specific test location. There is also information on whether a test was fully completed or not (up to 1989B), what type of test the person took, or if there was any type of rule violation (from 1990A). Since the information about the test type reveals certain diagnoses, this information is considered sensitive.
One of the few exceptions mentioned earlier occurred in the spring of 2020 when Covid-19 pandemic hit. For the first time, a test occasion was canceled. In the following years (2021-2022), the number of test occasions was increased to two tests each spring to reduce the spread of infection. Therefore, there is also a variable to distinguish these occasions.
The personal identification numbers of test participants are registered to ensure that the correct results are assigned to the correct test participants. Note that access to personal identification numbers is extremely limited and comes with many restrictions, as always when personal data is handled. Based on the personal identification number, information about the test participant's birth year, age range, and gender is generated.
In addition, there is also information about the test participants' educational level. However, it is important to note that the availability and range of education have undergone significant changes during the years that the SweSAT has been conducted. These changes have affected the coding of the educational variables and the periods during which the variables have been in use. It should also be noted that it is the test participants themselves who answered the questions regarding educational level at the time of the test.
Some variables have only been used for shorter periods, containing information about work experience (1977A-1989B), most recent education (1977A-1986B), further education (1982A-1986B), highest Swedish education (1987A-1989B), and whether one attended a high school line or program (1994B-1995B). In the spring of 1987 as well as the spring of 1990, a number of new variables were added to register this information. See the table below for a summary.
Here, the information regarding the test participants' results is presented, information concerning individual subtests, overarching test sections, or the entire test.
At the very first test occasion of the SweSAT, the test consisted of a total of six subtests: Vocabulary, Data sufficiency, Reading comprehension, Diagrams, tables, and maps, General orientation (discontinued after 1995B), and Study techniques (discontinued after 1991B). While some subtests have been discontinued, some have been retained, albeit with varying numbers of tasks in different versions of the test, and others have been added. In the spring of 1992, English reading comprehension was added, and during a major restructuring for the autumn test of 2011, Sentence completion, Mathematical problem-solving, and Quantitative comparisons were added. More information about the inclusion of subtests and the number of tasks can be found in the table below. The restructuring also meant that the test was divided into two overarching sections, a verbal and a quantitative section, each containing 80 tasks. Another change was that instead of administering each subtest individually, the tasks from the subtests were distributed across four blocks; two with quantitative tasks and two with verbal tasks.
Related to these subtests, there is information about the test participants' answers and scores on each subtest, raw scores and normed total scores, and since 2011B, raw scores and normed scores on the two test sections.
