This appendix shows the sample selection rules with respect to which programs were selected into the samples and which were rejected with the aid of a graphic (see below the text) which sketches out a number of possible cases.
As shown in the graphic, students B, E, and G are chosen into the sample:
Student B has one regular program during the first year. This program starts after the start date of the report cycle of the current year, and there are no other programs this year.
Student E has two regular programs during the first year, but the first ended “cleanly” before the beginning of the second, so it is considered a new program and is selected into our sample.
Student G has two regular programs during the first year. He/she is chosen because the later program starts after the start date of the report cycle of the first year, and it is clean (the earlier one has finished before the start of the later one).
Students (A, C, D, F, and H) are not chosen into the sample:
Student A has one regular program during the current year. He/she is not chosen, however, because this program starts before the start date of report cycle of the first year, so we did not observe its start and are unable to track it from that point.
Student C has two regular program records during the first year. He/she is not chosen because both programs start before the start date of report cycle of the current year.
Student D has two regular programs during the first year. He/she is not chosen because, although both programs start after the start date of report cycle of the first year, the two programs start simultaneously.
Student F has two regular programs during the first year. He/she is not chosen because although both programs start after the start date of the report cycle of the current year, the earlier one has not finished before the start of the later one.
Student H has two regular programs during the first year. He/she is not chosen because although the later program starts after the start date of the report cycle of the current year, the earlier program has not finished before the start of the later one.