The Scottish Government

« Previous | Contents | Next »

Listen

Chapter Seven Predicting which pupils are likely to work long hours

The importance of pupils' working hours

7.1 The impact of pupils' working hours on their school work and attainment is a concern for many stakeholders, especially teachers. A range of studies have examined this and concluded that the effect varies depending on the number of hours and the age and stage of the pupils concerned. Working long hours tends to have a negative impact (but the critical number of hours varies for different pupils) while working for a small number of hours may have a positive effect on attainment (Hobbs and McKechnie 1997; McKechnie and Hobbs, 2001; Payne, 2001; Stern and Briggs, 2001; McKechnie et al, 2002).

7.2 Given this concern about working hours, we wanted to identify which pupils are likely to work longer hours in their part-time job and therefore to be at risk of performing less well at school.

7.3 In chapter 6 we described pupils' working hours in respect of a number of factors. But as we have noted earlier (see p.45, chapter 4) while such analyses are useful in describing pupils' working hours they are limited because they deal with each factor separately. They do not show us the inter-relationship between the various factors or which ones are more important in predicting which pupils will work longer hours. Statistical modelling lets us consider all the factors at the same time and allows us to assess which factors are more influential than others. In this way we aim to identify the pupils who are most likely to work for longer hours in their part-time job.

Which pupils are likely to work long hours?

7.4 We modelled pupils' working hours to consider the various factors that might influence the number of hours that pupils are likely to work and to assess the impact of each when others are taken into account. ¹⁰ We carried out the modelling in a series of steps, introducing additional factors at each stage and omitting other factors that had proved to be non-significant. Here we present the final model of this process focussing on the significant findings (Table 7.1; see also Table 2, Appendix 2). It should be remembered that where any factor is significant, this is after taking account of the impact of the range of other factors.

Table 7.1: Current workers: predicting number of hours worked each week (linear regression)

The effect of type of job and employer

7.5 As we saw in earlier chapters, the research has used two job type classifications, the sixteen category system that pupils responded to in the survey and the five category system where we grouped these sixteen categories into five larger groupings: delivery, babysitting, retail, catering and miscellaneous. In considering the effect of type of job on working hours, we use the full job type classification. We take delivery work as the job type against which we compare the other jobs (ie the reference category), since, as we saw in chapter 5, pupils in delivery work have the lowest average working hours.

7.6 Table 7.1 shows that it is the type of job that pupils are employed in that is the key predictor of their working hours even after taking the other factors into account. Job type has a substantial influence on the number of hours, pupils employed in farm work were likely to work 9.9 hours longer each week than someone employed in delivery work. Working in a fast food outlet compared with delivery work increased working hours by 8.4 hours per week; being employed in a hotel/B&B added 6.7 hours each week while supermarket work increased pupils' likely working week by 6.4 hours.

7.7 Who pupils worked for also had a slight effect on their working hours. Pupils who were employed by their family were likely to work slightly longer hours compared to working for a non family employer; they were also liable to work longer hours if they were self employed (0.79 hours a week more in both cases).

The effect of background factors

7.8 As might be expected from the descriptive statistics already presented, school stage had a considerable effect on pupils' working hours, all other factors considered. The probability of working longer hours increased by stage, compared with current workers in S3, those in S4 were likely to work over an hour a week longer, S5s two hours a week more while pupils in S6 were likely to be employed for nearly three and a half hours more each week (+3.4 hours).

7.9 Gender had an effect but it was minor compared with the effect of type of job and school stage. Girls were likely to have a slightly shorter working week than boys (only 0.78 of an hour less each week).

7.10 After taking all the other factors into account, we found only a weak relationship between attainment and hours of work. Compared with those in the highest attainment group, those with 1-3 and 4-7 Credit SGs worked slightly longer but by less than an hour more.

7.11 The main difference in respect of social class and working hours concerns those whose father was in the 'missing' category: compared with current workers who came from a managerial/professional background, they were likely to work 2.4 hours a week longer. The difference between pupils from working class backgrounds and those from managerial/professional was minor (+0.6 hours for working class pupils).

7.12 Even after social class (and the other factors) were taken into account, family living arrangements had an independent effect on the intensity of pupils' working hours. Compared to pupils living with their mother and father, those pupils living with another relative were likely to work an additional 2.75 hours per week.

7.13 Pupils who classified themselves as being from 'other white backgrounds', ¹¹ or from Pakistani backgrounds were likely to work longer hours. Compared to Scottish pupils, both of these groups worked an additional 3 hours per week.

The effect of attitudinal factors

7.14 We reported in chapters 3 and 4 the link between having a current job and ever having truanted but it was also noted that the frequency of self-reported truanting was not strongly linked to having a current job. We found again in this analysis that truanting was significantly linked to hours worked. However, in contrast to the likelihood of being in a job or not, the frequency of truanting was important. Compared to pupils who reported that they never truant those pupils who truant for 'a lesson here and there' were likely to work an additional 0.9 hours per week while those who truanted for days at a time worked an additional 2.49 hours per week. We should be cautious in interpreting this since we cannot assume that they truant in order to work. As we have already reported, when pupils who had truanted were asked if they had engaged in part-time work when they were absent, 86% responded that they had not done so.

7.15 Few of the other attitudinal factors were linked to the number of hours worked. However, pupils who feel that school is not preparing them for life after school were more likely to commit more time to their jobs. In this case the impact is slightly less than for truanting, adding only 0.44 hours per week.

7.16 Pupils self assessment of their enterprising attitudes was also found to have an independent effect on the likely number of hours that pupils would work. Those pupils with the highest enterprising scores were likely to work slightly longer hours each week, though the difference is small (0.52 hours).

The effect of career related factors

7.17 We noted in the previous chapter that pupils with a definite career focus were more likely to be currently employed. Career focus is also linked to the number of hours worked, after controlling for the other factors. Those pupils who had a definite career focus had a slightly greater likelihood of working longer hours although at +0.29 hours per week the impact is not large.

7.18 Post school plans had a separate impact on working hours. Compared to pupils who were intending to enter HE those pupils who were planning to be self employed after leaving school were likely to work an additional 1.4 hours per week. However, the greatest impact on hours is associated with pupils who responded 'don't know' when asked about their post school plans. These pupils were likely to work an additional 2.12 hours per week, compared to pupils who planned to enter HE.

The effect of voluntary work

7.19 So far we have focused on those factors which predict a greater intensity of work, in fact we found only one case where any factor predicted a reduced number of working hours - this is involvement in voluntary work. Current workers who did voluntary work were likely to work for fewer hours per week than other working pupils who were not also engaged in voluntary work. The impact predicted from this model is that such activity leads to pupils working 1.3 hours per week less than pupils who do not do voluntary work. We earlier found that participating in voluntary work does not predict whether a pupil has a current job or not but if pupils have a job, then it does appear to reduce the number of hours worked.

Non significant factors

Living in a rural area

7.20 It is also worth noting the factors which we found to be non-significant in the final model or in earlier stages of the modelling. These include location (large urban, other urban and rural). In the descriptive statistics in chapter 5, we saw that pupils in rural areas had the highest weekly average working hours. . . In the final stage of the modelling, however, living in a rural area does not quite reach statistical significance. ¹² In earlier stages of the modelling, living in a rural areas did have a significant effect on working hours even after controlling for various factors including the type of job that pupils did and their stage of schooling. It was when we added the factors relating to career plans and focus that the effect of location was reduced and no longer reached the level of significance that we report in this research.

Other time commitments

7.21 Although we found that participation in voluntary work is likely to reduce working hours, none of the other time commitment factors had an impact on likely working hours. We examined the effect of the time that pupils who were currently working spent on homework, on housework and on care duties, the extent of their participation in various out of school activities and also whether they were doing any unpaid work. None of these factors influenced their likely working hours.

7.22 Other non significant factors include: disability; mother's current activity; work experience; work-related learning inputs; enterprise education inputs; career inputs; and most of the factors relating to attitudes to school.

Predicting the type of job that pupils will do

7.23 Given the importance of the type of job in determining pupils' likely working hours, we extended the analysis to consider pupils' probability of being employed in each of the main categories of employment. In the research we have used two job type classifications: a detailed one comprising sixteen different job types and a condensed five category system where similar jobs are grouped together in one category (delivery, babysitting, retail, catering and miscellaneous). It was not possible to use the detailed classification system in this analysis because of small numbers in some categories so we used the condensed groupings but excluded babysitting due to the relatively small number of pupils involved in this activity.

7.24 There is very substantial variation across the job types in the factors that predict the likelihood of employment but no clear pattern is evident and it is difficult to draw any specific conclusions. We therefore do not report the findings here but the results of the statistical modelling for each of the job types are given in Tables 3-6, Appendix 2. It is may be that the four large job categories we used in the analysis are too general to enable us to identify particular patterns; to do this we would need to use the more detailed sixteen job type classification to be able to but the number of cases are insufficient in some categories to permit this analysis.

Overview

7.25 The impact of pupils' working hours on their school work and attainment is a matter of concern and one that is pertinent to the question of the desirability of recognising pupils' part-time employment. Existing research demonstrates that the number of hours worked is critical and so we examined the factors associated with hours of work and have identified certain factors ones that mean that pupils are more likely to work longer hours:

• the type of job that pupils do is the most important factor in predicting their working hours. Pupils who work in farming, fast food outlets, hotels/B &Bs and supermarkets are likely to work the longest hours (+ 6 -10 hours for these jobs).

7.26 There are a number of other factors that predict longer working hours and which should also be considered when trying to identify which pupils are more likely to work longer hours:

• by school stage and especially by the S6 stage (pupils in the upper school and especially S6s pupils: +3.4 hours a week)
• ethnicity (pupils from 'other white backgrounds' and from a Pakistani background: +3.1 hours)
• living arrangements (where pupils live with other relatives rather than their mother and father: +2.7 hours)
• social class (father in the missing category: +2.4 hours)
• truancy (truant for 'days at a time': +2.49 hours). But it is notable that most of the factors relating to attitudes to school made no significant difference to working hours
• post school plans (don't know: +2.12; own business +1.48).

7.27 While attainment and the extent to which pupils judged themselves to be enterprising were each significant in predicting hours of work, they made only a minor difference, less than an hour each week. It seems that participation in voluntary work, while it does not reduce the likelihood of pupils' having a part-time job, is associated with working fewer hours. It might be noted that none of the other out of school activities that pupils engage in had an effect on working hours.

7.28 Since the type of job worked in emerged as the critical factor in predicting their likely working hours, we extended the analysis to consider pupils' probability of being employed in a particular job category. However, we were only able to consider this in respect of four main categories of jobs rather than for the more detailed classification of 16 job types. The results of this analysis were inconclusive. It is evident that it is necessary to look in more detail at the specific job types to identify the factors that influence the type of job that pupils will have rather than using the broader job categories; the data, however, did not permit such an approach.

« Previous | Contents | Next »