The First Year Success Program: Do Bridging Courses Improve Success and Retention at University for Equity Students?

The intervention is the First Year Success Program (FYSP). There are three main components of the FYSP: 1) academic bridging courses, 2) an introductory “HeadStart” workshop and 3) peer mentoring. The HeadStart workshop is the first of the FYSP sub-programs and takes place in mid-January. It provides students with important details and guides for starting university. The peer mentoring program runs throughout the first semester, where students will have access to a peer who is an existing UTS student with whom they can ask questions and seek guidance from. The academic bridging courses are delivered in-person and on campus. These courses start from late January to mid-February, prior to the semester commencing. There are 5 separate bridging courses: mathematics preparation, mathematics, physics, chemistry and biology. The courses are designed to provide students with foundational skills in the areas needed for their degree, which is assumed prior knowledge. The courses run between 8 – 10 days and each daily session is on average 3 hours, resulting in contact hours of around 30 hours per course (with slight variations across each course). In 2025, the cost is $396.00 per course and are available to all students at this price. However, students in the FYSP are able to access the courses for free.

2026-01-13

2026-02-28

Standardised student grades, re-enrolment/attrition rate

Our main outcomes of interest are standardized student grades in their first-year subjects and whether they are re-enrolled in the 2nd semester of their first year or have left or paused their studies. Information on student marks and enrolment are directly measured through the UTS administrative system. Grades are measured at the subject level as a raw score on a 0 to 100 scale, and we transform this into a standardized measure with a mean of zero and a standard deviation of one. This is done by subtracting the mean of all grades from each student’s score for each subject and then dividing by the standard deviation of the grades. The resulting outcome will be standardised student grades at the subject level. Enrolment in the 2nd semester will be a binary indicator equal to zero if the student re-enrols and one if the student has left or paused their degree. The mean of this variable is the attrition rate or probability that a student fails to continue their studies.

success rate (weighted and unweighted), weighted standardised grades

We also use success rate as a secondary outcome, measured as the number of subjects passed divided by the number of subjects taken. To take into account the relative importance of different subjects we also calculate the weighted success rate which is measured by the sum of all the credit points from subjects successfully passed divided by the total sum of credit points attempted that semester. For weighted measure of standardised grades, we will first calculate the standardised scores using the mean and standard deviation of all students. Then, we will apply subject-level weights to the standardised scores which will weight each score according to how many credit-points the subject is worth out of the total amount of credit points taken in that semester.

We implement a Randomized Controlled Trial. Participants in our trial will be randomly allocated into treatment and control groups. We use block randomisation to achieve balanced group sizes within each equity group and faculty. The treatment group will receive an offer to participate in the FYSP whilst the control group will not receive anything. This means there will be non-compliance from both treatment and control, as not everyone in the treatment will attend and some of the control group may attend the bridging courses (which can be accessed for a fee). In the offer letter and email, the treatment group will be encouraged to participate in the 3 programs. For the bridging courses we inform the treated students that they can attend the course, which is normally priced at $396, for free. We also inform these students that they will be paid if they attend the bridging courses. The payment is calculated and sent via email for an e-voucher daily. The maximum payment a student can receive is $100. This payment is framed as payment cover any personal costs associated with attendance, rather as an incentive for attending. The framing of the payment is consistent with other experimental evidence (Leuven et al., 2010) which shows that financial incentives can crowd out student’s intrinsic motivation to attend, leading to lower attendance. Put simply, students will be more likely to attend because they are intrinsically motivated to do so instead of being compensated for it. These initial round of communications to the students in the treatment group will also include small informative nudges, designed to highlight the potential difficulties of transitioning to university and how attending this program may help them.

Not available

We generate random numbers in Stata and allocate treatment status accordingly

We use stratified block randomisation where the strata are faculty and equity group and the blocks are offer rounds but randomisation is at the individual level within those stratified blocks.

No

0

3,500

1,750

The introduction of the opt-in mechanism alters standard power calculations and introduces a broader range of possibilities. The final sample size is now determined by how many students agree to participate in the study and this has implications for statistical power. In presenting the power calculations we show a variety of scenarios for both the number of students opting in to the study and the number of students who participate in the study. In determining the minimum detectable effect (MDE), we follow the standard calculations presented in equation 1 in the appendix. We then display all the parameter values in table 1 in the appendix. The MDE is a function of the level of statistical power, conventionally set at 80%. The level of statistical significance is set to 5%. The calculation assumes that 50% of the sample allocated to the treatment group (which maximises power). Student marks will be standardised (i.e. to have a mean of 0 and variance of 1). Attrition will be measured as a binary outcome (i.e. equal to one if a student leaves their studies and zero otherwise), the variance is estimated based on the assumed mean of 20% attrition. The corresponding MDE values for different scenarios of opt-in and take-up rates are presented in table 2 and 3 for the attrition rate and standardised grades respectively. It is evident that high levels of recruitment and program take-up provide the largest amount of statistical power (these are the scenarios where the MDE is the lowest). It is difficult to ascribe likelihoods over the various scenarios shown in each of the tables. However, we believe that our recruitment rate is likely to lie between 50% – 70% and our take-up rate between 30% – 50%. This yields the following range of MDEs for attrition between 9.14 and 18.03 percentage points and for standardised grades between 0.229 and 0.451 standard deviations.

Trial Analysis Plan