Logistic Regression Analysis



Main Non-Progression Models

This section presents the results of regression analysis that models non-progression rates controlling for a wide range of student and institutional level characteristics.

Comparison of unadjusted rates of non-progression may be confounded by differences in individual characteristics and or institutional factors. To further understand the key drivers of non-progression, student characteristics such as leaving certificate points, age, gender, socio-economic background, domicile, school type, and entry basis are included as independent variables in the regression models. In terms of the institutional factors, institute, NFQ level and field of study are also included in the models. This chapter attempts to disentangle the effects of these variables and model non-progression rates based on like-for-like students.

Non-progression can be modelled in two ways. The first is the unadjusted (before controls) non-progression rate, which is the same as the headline rates reported elsewhere in this report. The adjusted (after controls) estimates compare like-for-like students to isolate the impact of the selected variable on non-progression. The default variable selected in Figure 1 is institute. These results compare headline rates of non-progression with like-for-like students who attend different institutions but who study the same subject, at the same NFQ level, are the same age and gender, received the same Leaving Certificate points, attended the same type of secondary school, entered higher education via the same route and are from the same socio-economic background. Comparisons of adjusted rates of non-progression – over time or across student/institutional characteristics – provide a more nuanced understanding of non-progression rates in Ireland.

Figure 1 - Non-Progression: Before and After Controls

Dashboard 5

var divElement = document.getElementById(‘viz1648217632703’); var vizElement = divElement.getElementsByTagName(‘object’)[0]; if ( divElement.offsetWidth > 800 ) { vizElement.style.width=’100%’;vizElement.style.height=(divElement.offsetWidth*0.75)+’px’;} else if ( divElement.offsetWidth > 500 ) { vizElement.style.width=’100%’;vizElement.style.height=(divElement.offsetWidth*0.75)+’px’;} else { vizElement.style.width=’100%’;vizElement.style.height=’727px’;} var scriptElement = document.createElement(‘script’); scriptElement.src = ‘https://public.tableau.com/javascripts/api/viz_v1.js’; vizElement.parentNode.insertBefore(scriptElement, vizElement);

Figure 1 compares the unadjusted and adjusted rates of non-progression across HEIs for new entrants in the academic year 2019/20. Significant variation in non-progression rates can be observed across HEIs before controls are applied. For example, the headline rate of non-progression in Dundalk IT in 2019/20 was 16%, followed by Limerick IT, IT Carlow, IT Sligo at 15%, respectively. At the other end of the scale, the non-progression rate in the Royal College of Surgeons Ireland (RCSI) and St Angela’s College was just 3% in the same year. Comparing like-for-like students reduces the variance in non-progression rates across HEIs. For example, the predicted non-progression rate in Dundalk IT and Limerick IT reduced to 11%, while in Sligo IT the predicted non-progression rate was 10%, after controls were included. On the other hand, the predicted non-progression rate in the Royal College of Surgeons and St. Angela’s College is marginally higher when like-for-like students are compared1.

Using the available drop-down filters in Figure 1, users can compare adjusted and unadjusted non-progression rates across years and by variable. For example, non-progression rates across Leaving Certificate Points Bands can be explored by selecting ‘Leaving Certificate Points’. Both the adjusted and unadjusted estimates illustrate a strong negative relationship between Leaving Certificate Points and non-progression rates. For example, the average non-progression rate for students with less than 250 Leaving Certificate Points is above 20%, which is evident for both unadjusted and adjusted estimates. Conversely, students with higher Leaving Certificate Points (greater than 500 points) have significantly lower non-progression rates, before and after controls are applied.


 

Non-Progression by Socio-Economic Background, Gender and Leaving Certificate Points

Detailed analysis looking specifically at the relationship between Deprivation Index Scores, Leaving Certificate Points, and Gender provides further insight into the factors affecting non-progression rates in Ireland. To properly understand these relationships, all missing observations and unknowns are excluded. Moreover, this analysis includes data for the class of 2019/20 only. In this context, the sample is reduced to 28,392, resulting in an average non-progression rate of 8%. Results in this section are therefore not directly comparable to figures provided elsewhere in this report.

Deprivation Index Scores provide a measure for the relative affluence or deprivation of the student population1. Before controls are applied, a negative relationship between relative affluence and non-progression can be observed (Figure 2). For example, non-progression rates were twice as high among students from the most disadvantaged backgrounds (12%) compared to the most affluent students (6%). However, after controlling for Leaving Certificate Points, no clear relationship between Deprivation Index Scores and non-progression can be observed.

Disparities in Leaving Certificate Points between disadvantaged and affluent students largely explains the difference in non-progression rates between these two groups. This suggests that non-progression rates are lower for affluent students because, on average, they have higher Leaving Certificate Points than disadvantaged students. Conversely, students from lower socio-economic backgrounds typically have lower Leaving Certificate Points, which explains why, on average, these groups of students have higher non-progression rates. This does not to imply that relatively high non-progression rates among disadvantaged students is not an issue per se, rather it serves to highlight that the factors affecting non-progression are more nuanced than the headline rates may indicate.

Figure 2 - Non-Progression: Deprivation Index Scores

var divElement = document.getElementById(‘viz1647000898948’); var vizElement = divElement.getElementsByTagName(‘object’)[0]; if ( divElement.offsetWidth > 800 ) { vizElement.style.width=’100%’;vizElement.style.height=(divElement.offsetWidth*0.75)+’px’;} else if ( divElement.offsetWidth > 500 ) { vizElement.style.width=’100%’;vizElement.style.height=(divElement.offsetWidth*0.75)+’px’;} else { vizElement.style.width=’100%’;vizElement.style.height=’727px’;} var scriptElement = document.createElement(‘script’); scriptElement.src = ‘https://public.tableau.com/javascripts/api/viz_v1.js’; vizElement.parentNode.insertBefore(scriptElement, vizElement);

Non-progression rates by gender have been well documented in previous HEA analysis. A common theme has emerged, showing that males typically have higher rates of non-progression than females. Findings in this report, as shown in Chapter 2, are reflective of these stylised facts. It is noteworthy that disparities between males and females, in terms of non-progression rates, remain despite comparing like-for-like students, albeit the gap is reduced (see Figure 1 above).

Figure 3 illustrates how average non-progression rates vary by gender across the Leaving Certificate Points distribution. Specifically, Figure 3 highlights that non-progression rates vary significantly by gender, depending on prior academic ability. For example, little variation in non-progression rates between males and females is observed for those with more than 450 Leaving Certificate Points. However, males with less than 350 Leaving Certificate Points have higher non progression rates than females with similar Leaving Certificate Points. The gap is largest for students at the lower end of the Points distribution, with males almost twice as likely to not progress (compared to females) where they enter higher education with 200 leaving certificate points or less. Similar findings are observed when like-for-like students are compared, albeit the non-progression gap between males and females with lower leaving certificate points is marginally smaller.

Earlier in this chapter, Figure 1 highlighted that the predicted non-progression gap between males and females was just 2 percentage points when like-for-like students are compared. Figure 3 however, shows that the probability of progressing varies significantly for males and females depending on Leaving Certificate Points. These results reveal the significant heterogeneity in non-progression rates by gender across the Leaving Certificate Points distribution.

Figure 3 - Non-Progression: Gender and Leaving Certificate Points

var divElement = document.getElementById(‘viz1646921675809’); var vizElement = divElement.getElementsByTagName(‘object’)[0]; if ( divElement.offsetWidth > 800 ) { vizElement.style.width=’100%’;vizElement.style.height=(divElement.offsetWidth*0.75)+’px’;} else if ( divElement.offsetWidth > 500 ) { vizElement.style.width=’100%’;vizElement.style.height=(divElement.offsetWidth*0.75)+’px’;} else { vizElement.style.width=’100%’;vizElement.style.height=’727px’;} var scriptElement = document.createElement(‘script’); scriptElement.src = ‘https://public.tableau.com/javascripts/api/viz_v1.js’; vizElement.parentNode.insertBefore(scriptElement, vizElement);

[1] Due to poor data coverage of student level information for the RCSI, adjusted non-progression rates should be treated with caution. As a consequence, little variation in before and after controls estimates are observed for the RCSI.

[2] Deprivation Index Scores range from approximately -40 (most disadvantaged) to +40 (most affluent). Due to small numbers of observations at both tails of the distribution in this sample, non progression rates are estimated on a scale of -20 to +20. More information on Deprivation Index Scores can be found here.

Our use of cookies

We use necessary cookies to make our site work. We'd also like to set optional analytics cookies to help us improve it. We won't set these optional cookies unless you enable them. Using this tool will set a cookie on your device to remember your preferences.

For more detailed information about the cookies we use, see our Privacy Policy page


Necessary cookies

Necessary cookies enable core functionality such as security, network management, and accessibility. You may disable these by changing your browser settings, but this may affect how the website functions.


Analytics cookies

We'd like to set Google Analytics cookies to help us to improve our website by collecting and reporting information on how you use it. The cookies collect information in a way that does not directly identify anyone.