Summary

This analysis builds on previous studies on Modern Apprenticeships and industry training using each programme's administrative dataset to determine if Modern Apprenticeship's additional supports and structures are effective tools to ensure engagement and achievement in formalised industry training programmes for younger people.

Key findings:

1. When other factors are controlled for, there seems to be a premium in the Modern Apprenticeships model over normal industry training, manifested in programme completion rates. On aggregate, learners engaged in Modern Apprenticeships are more likely to complete their programme than equivalent learners in industry training.
2. This is not true in all industries. Modern Apprentices completed their programmes at higher rates than equivalent industry trainees in just over half of the matched ITOs. There may be industry-specific factors and/or programme administration factors that mean that the Modern Apprenticeship model works better to provide a completion premium in some industries than it does in others.
3. There does not appear to be any premium on completion between industry training and Modern Apprenticeships when combined with the ethnicity of the learner. When other factors are adjusted for, each ethnic group performs relatively similarly between the two programmes.
4. When other factors are adjusted for, there appears to be no difference in coordination effects based on the identity of the coordinator (ITO coordinators compared to non-ITO coordinators) in Modern Apprenticeships in respect to likelihood of learners to complete their programmes.  Observed higher completion rates for non-ITO coordination services may be a function of differences in brokerage practices, specifically recruitment criteria, between the two.

The Modern Apprenticeships programme was introduced nation-wide in 2001 to address participation problems in workplace industry training by young people. It is aimed at 15 to 21 year olds wishing to participate in formalised workplace-based training, and is intended to lead to national qualifications. It is based on the traditional industry training arrangements, but differs in two important ways.

Modern Apprenticeships involves additional support for both the apprentice and their employer. Modern Apprenticeships coordinators both act as apprenticeships brokers, arranging for job placements for young people, and provide ongoing support and assistance to both employers and apprentices.

Both programmes are administered by the Tertiary Education Commission (TEC), and both are intended to lead to attainment of national qualifications. Modern Apprenticeships qualifications each contain a larger quantum of learning than is usual for non-targeted industry training. An average of 120 credits in total is required across a period of approximately four years for each Modern Apprenticeships learner. Industry training programmes are often smaller, with some programmes consisting of only 40 credits per learner, and they are taken over varying time periods, depending on the requirements of each participant and associated workplace. They do not usually involve placement in a new job, as participants are already involved in employment before starting training.

Industry Training Organisations (ITOs) play a standard-setting and an assessment-arranging role for both programmes. ITOs provide sole Modern Apprenticeships coordination services in some industries. In others, non-ITO organisations provide coordination services, while in others, coordination services are provided by both ITO and non-ITO organisations.

There are several possible methods to measure success in industry training and Modern Apprenticeships. Analyses of programme variables have used programme completion as an indicator of the outcome of each training event. Programme completion is used in this analysis as a quality indicator, and the likelihood of a learner completing a least one programme in Modern Apprenticeships and industry training is compared.

Despite the difference in the quantum of learning required of the two programmes, it is possible to compare them by controlling for the differences between them, using a statistical method called logistic regression and using matched cohorts of learners. Using a matched cohort, logistic regression controls for the effects of learner and programme characteristics on outcomes, and allows us to make inferences about how outcomes change as the value of one explanatory variable changes.

In assessing the differences (if any) between the two programmes, we might want to know the answers to the following questions:

1. Do Modern Apprentices complete their programmes at similar rates to industry trainees?
2. Do non-European Modern Apprentices do better or worse than in industry training?
3. Do Modern Apprenticeships learners' chances of success differ according to who provides coordination services?

Apart from the coordination services, there are some fundamental differences between the two programmes that need to be controlled for before a meaningful comparison between the two can be made. For example, previous analyses have identified that duration of learning is positively correlated with completion in Modern Apprenticeship programmes, and negatively correlated with completion in industry training programmes. That is, it seems that the longer a period of training has been at exit in industry training for each individual, the less likely it is to be as a completed programme. But the reverse applies in Modern Apprenticeships – the longer the training period, the more likely the programme is to end in a completion.

The corollary is that after short durations of learning industry training learners complete their programmes at higher rates than Modern Apprentices. This is in part a consequence of the shorter (lower credit quantum) programmes in industry training. Apprentices who leave programmes early are more likely to be non-completers than leavers in the less rigid, variable-duration industry training. Modern Apprentices do complete their programmes at higher rates than equivalent industry trainees on aggregate, but over longer periods of time.

Previous analyses explored the probability that learners in industry training and Modern Apprenticeships complete their programme. Mahoney (2009a) found that the predicted and observed probability of a learner competing their programme in industry training is 33 percent. An estimated 35 percent of learners starting industry training for the first time in 2003 completed at least one programme within five years.

Modern Apprenticeships completion rates improved from 32 percent to 40 percent between 5 and 6 years after commencing study, and 42 percent of 2002 Modern Apprentices starters had completed at least one programme after 7 years. This implies that Modern Apprentices often take a long time to finish their qualifications – longer than the number of years envisaged when the programme was originally devised.

This study controls for the differences in programme credit load between industry training and Modern Apprenticeships, while ignoring specific programme duration, to see if there is a difference between industry training and Modern Apprenticeships in respect to programme completion when both are put on the same footing.

Within Modern Apprenticeships, the observed completion rates seem to indicate some differences between ITO and non-ITO coordination services, as well as differences between ethnic groups. However, both of these studies found that various other variables, such as industry, previous qualification, rate of study and ethnic group (for example) are also strongly associated with completion. Mahoney (2009b) found that in some situations non-ITO coordination seems to result in higher completion rates than ITO coordination services. This analysis tests whether observed differences in completion rates (for example those published on Education Counts) are due to differences in services provided between them, or due to clusters of combinations of other factors associated with success.

Similar issues exist between industry training and Modern Apprenticeships where differences in the performance of some groups over others are observed. For instance, observed cohort completion rates indicate that Pasifika learners do better in industry training than Modern Apprenticeships. Is this a consequence of one programme suiting Pasifika learners more, or a consequence of clustering of  variables associated with success?

This study attempts to control for confounding factors to answer our three research questions. It follows the progress of an age and industry-matched group of learners who are identified as starting industry training or Modern Apprenticeships at around the same time, doing programmes with similar credit loads in matched fields of study, and examines the outcome of their learning after a maximum period of time (8 years) have passed. It controls for the varying durations in the two programmes by taking a cohort approach, thereby ignoring varying durations between and within the programmes.

Other variables are controlled for using logistic regression, such as:

• the programme credit load
• the rate of study
• the prior qualifications of the learner
• learner age at entry
• learner geographic location
• learner gender and ethnic group
• NQF level of programme
• Start year (2002, 2003, 2004 or 2005)
• the ITO administering their programme (a proxy for the industry they are working in)
• an identifier showing whether learners are in industry training or Modern Apprenticeships, and if Modern Apprenticeships, whether their coordinator is or is not an ITO.

Two statistical models are used to answer the three main research questions: Model 1 compares matched industry training and Modern Apprenticeships training, to answer questions 1 and 2. Model 2 compares outcomes within Modern Apprenticeships to answer question 3. Both models use the same cohort of learners identified below.

A brief discussion of the meaning of the results of the modelling is included in the conclusion section of this paper.