An analysis of the work schedules of Great Barrier Reef pilots

Results and Discussion

Work assignments were clearly identifiable from the work schedule files and were measured from the date and time of embarkation to the date and time of disembarkation (i.e. time on the ship). At this stage of the analysis no attempt has been made to split time on the ship into work on the bridge and rest; on tour logs will provide this specific information.

5.1 Number of Work Assignments per Tour

For each of the pilotage companies across the three 6-month periods, the average number of work assignments per tour is shown in Figure 5.0. Between company and time period differences from the analysis of this data are summarised in Table 5.0. The number of work assignments per tour differed significantly between the three companies. The greatest number of work assignments per tour were performed by Company B pilots (5.21) followed by Company A (4.59) and then Company C (1.59). The fewer number of assignments undertaken by Company C relates to the single operational region of this company.

By presenting the number of work assignments per tour as a series of frequency distributions (Figure 5.1a and Figure 5.1b), a clearer indication of the work patterns of Great Barrier Reef pilots can be obtained. For both Companies A and B, the frequency curves were fairly similar with most pilots performing 8 or fewer work assignments per tour. However, there were a number of situations for both companies in which considerably more work assignments were undertaken during a single tour, with some tours involving 14 or more work assignments. While unusual circumstances may have prevailed during these tours, the heightened risk of fatigue which exists in situations of extreme workload (Iskra-Golec et al. 1996; Rosa et al. 1989; Spurgeon et al. 1997) may justify closer monitoring of the number of work assignments being performed during tours.

When examining changes in the average number of work assignments per tour across the 18 months, a significantly greater number of work assignments (4.19) were performed in the second time period (Table 5.0). This trend is also evident on the frequency distributions (Figures 5.1a and Figures 5.1b), as most of the tours involving greater than average workloads were undertaken during the middle 6-month period. The same temporal pattern was observed for each company, thereby suggesting that the greater number of assignments in the second six months was a reflection of the dynamics in shipping volumes and available work. Personal communication with Pilot Advisory Group members (4 August, 1997) substantiates this suggestion.

The finding of significant variations in the amount of pilotage work across time highlights the unpredictability associated with marine pilotage work. Situations such as industrial disputation in mining areas in central and northern Queensland and the economic state of neighbouring Asian countries contribute to work fluctuations, as do seasonal variations in imports and exports (Personal Communication - Iain Steverson, 14 May, 1998). As a consequence, pilots are never really sure how much work will be available in the future. This uncertainty more than likely contributes to the stress experienced by Great Barrier Reef pilots with regards to financial income and job security (Personal Communication - Pilot Advisory Group, 4 August, 1997). It also makes the jobs of the pilotage company administrators more difficult when trying to ensure work is evenly distributed among the pilots and in the development of fatigue management plans (Personal Communication - Iain Steverson, 14 May, 1998).

Figure 5.0 Mean number of work assignments per tour, by Company and period (1 Jan 1996 - 30 June 1997)

Note: Data analysis based on n values shown in Methods (Section 3.2).

Table 5.0 Analysis of the mean number of work assignments per tour (1 Jan 1996 - 30 June 1997) (1)

1. Results of full two-way analysis of variance (ANOVA) model including company and period effects. p < 0.01 considered statistically significant.
2. Results of Tukey’s Studentised Range Test for post-hoc differences (Type I Experimental Error Rate = .01)

Company (Main Effect). This term measures whether or not there is a difference between companies averaged across the three time periods.

Period (Main Effect). This term measures whether or not there is a difference between the periods averaged across the companies.

Company * Period Interaction. This term measures whether or not the difference between the companies is the same in each period. Equally, it measures if the difference between periods is the same for each company. When the interaction is statistically significant there is little interest in the main effects. In this case averages over one of the factors ignore important variations.

Figure 5.1a Frequency distribution - Number of work assignments per tour by period - Company A

Figure 5.1b Frequency distribution - Number of work assignments per tour per period - Company B

5.2 Number of Work Assignments per Pilot

Presented in Figure 5.2 are the mean number of work assignments per pilot per six month period, for each company separately. Table 5.1 summarises the results of the analysis of this data in terms of company and period differences. The three companies differed significantly in the mean number of assignments performed per pilot; Company B pilots performed the greatest number of assignments (28.6), followed by Company A pilots (21.8) and then Company C pilots (7.4). When the two pilot groups with similar operational regions are compared (i.e. Companies A and B), these results indicate pilots working for Company B performed approximately 31% more work assignments than their colleagues in Company A. Such a finding is consistent with previous results indicating that Company B pilots spent more time on tours and had shorter breaks between tours during each 6 month period (Section 4). The smaller number of assignments per pilot performed by Company C reflects the single operational region of this company.

That pilots from Company B are undertaking significantly more work assignments per 6 month period compared with Company A personnel may be a reflection of increased commercial pressures generated by the privatisation of pilotage in the Great Barrier Reef - Torres Strait region. Competition between pilotage companies for work is now an integral part of Great Barrier Reef marine pilotage, with each company trying to secure enough work to make their operation economically viable. It is therefore possible that over the 18 month analysis period, Company B may have secured more work than Company A. The increased job security and financial benefits of this additional work would have undoubtedly been welcomed by Company B pilots. However, the increased risk of fatigue development associated with greater workloads and potential impact this may have on safety standards (Spurgeon et al. 1997) raises concern over the appropriateness of competition between pilot companies.

Concern over the issue of competition in pilotage operations was raised in a paper presented at a recent International Pilotage Conference held in Brisbane. The author indicated that competition is incompatible with compulsory pilotage; it is bad for the public, for the shipping industry and the pilotage profession. Competition may compromise safety, particularly when pilots are expected to exercise independent judgment and resist pressures, which are inconsistent with the interests of safety. Commercial pilotage operations may also lead to the provision of discriminatory service, and may be economically inefficient requiring duplication of many items of expense such as pilot launches and dispatch services (Sparks, 1998).

Post hoc analysis of changes across time (Table 5.1) revealed that significantly more work assignments were performed by pilots during the second 6-month period (21.4), as compared to the first period (14.8). Results for the third time period were inconclusive. These changes were similar for each of the pilotage companies and support the previous finding indicating increased work availability during the second time period. Consultation with pilotage company personnel indicated that the increase in shipping traffic during the period between July 1 and December 31 is an annual occurrence related to the sugar cane season in northern Queensland. However, as other factors also influence work availability such as industrial disputes and the economic state of neighbouring Asian countries, it is somewhat difficult to predict how much work will be available during any one period (Personal communication – Iain Steverson, June 10 1998).

By presenting the number of work assignments per pilot as frequency distributions (Figure 5.3a and Figure 5.3b), it is apparent that in both Companies A and B there were a number of pilots performing considerably greater workloads than what the company average figures depict. This was especially the case during the busy second 6 month period and for personnel from Company B. Given the increased fatigue potential of greater workloads (Iskra-Golec et al. 1996; Rosa et al. 1989; Spurgeon et al. 1997), the finding that there were personnel in both companies undertaking in excess of 35 work assignments per 6-months is disconcerting and implies that closer monitoring of the situations influencing the number of work assignments undertaken by pilots may be required.

Figure 5.2 Mean number of work assignments per pilot, by Company and period (1 Jan 1996 -30 June 1997)

Note: Data analysis based on n values shown in Methods (Section 3.2).

Table 5.1 Analysis of the mean number of work assignments per pilot (1 Jan 1996 - 30 June 1997) (1)

1. Results of full two-way analysis of variance (ANOVA) model including company and period effects. p < 0.01 considered statistically significant.
2. Results of Tukey’s Studentised Range Test for post-hoc differences (Type I Experimental Error Rate = .01)

Company (Main Effect). This term measures whether or not there is a difference between companies averaged across the three time periods.

Period (Main Effect). This term measures whether or not there is a difference between the periods averaged across the companies.

Company * Period Interaction. This term measures whether or not the difference between the companies is the same in each period. Equally, it measures if the difference between periods is the same for each company. When the interaction is statistically significant there is little interest in the main effects. In this case averages over one of the factors ignore important variations.

Figure 5.3a Frequency distribution - Number of work assignments per pilot per period - Company A

Figure 5.3b Frequency distribution - Number of work assignments per pilot per period - Company B

5.3 Number of Work Assignments per Pilot – Inner Route

Given the overall priority of the study to identify factors possibly associated with fatigue, a breakdown of work assignments by shipping route was undertaken. The section of the Inner Route which stretches between Cairns and Cape York, is a region of compulsory pilotage. Some characteristics of the Inner Route such as the longer duration of pilotage (up to 90 continuous hours) in combination with hazardous shipping conditions (tides and under keel clearance) have the potential to substantially increase stress and fatigue levels (Personal Communication - Pilot Advisory Group, 4 August, 1997).

Figure 5.4 shows time period and company results for the mean number of work assignments per pilot on the Inner Route. Results of the analysis of this data in terms of company and time period differences are shown in Table 5.2. Company C does not operate in this region and were therefore excluded from the present analysis. The number of work assignments undertaken on this shipping route was significantly greater for Company B pilots (23.6) than their colleagues in Company A (15.6). These figures reflect a 51% greater workload on the Inner Route for pilots of Company B. These results are also depicted by the frequency distributions (Figure 5.5a and Figure 5.5b) which show that substantially more pilots from Company B performed 25 or more assignments per 6-months on this shipping route, compared with Company A pilots.

In terms of changes across time in the number of assignments performed on the Inner route, significant differences existed. An average of 22.7 Inner Route passages were performed per pilot during the second 6-month period, as compared to 19 passages during the third 6-months and 16.4 during the first 6-months. These findings support previous results indicating increased work availability during the middle 6-month period of the analysis. The same temporal patterns in the number of assignments were observed in both companies.

Figure 5.4 Mean number of work assignments per pilot on the Inner Route by Company and period (1 Jan 1996 - 30 June 1997)

Note: Data analysis based on n values shown in Methods (Section 3.2).

Table 5.2 Analysis of the mean number of work assignments per pilot on the Inner Route (1 Jan 1996 - 30 June 1997) (1)

1. Results of full two-way analysis of variance (ANOVA) model including company and period effects. p < 0.01 considered statistically significant.
2. Results of Tukey’s Studentised Range Test for post-hoc differences (Type I Experimental Error Rate = .01).

Company (Main Effect). This term measures whether or not there is a difference between companies averaged across the three time periods.

Period (Main Effect). This term measures whether or not there is a difference between the periods averaged across the companies.

Company * Period Interaction. This term measures whether or not the difference between the companies is the same in each period. Equally, it measures if the difference between periods is the same for each company. When the interaction is statistically significant there is little interest in the main effects. In this case averages over one of the factors ignore important variations.

Figure 5.5a Frequency distribution - Number of work assignments per pilot per period on the Inner Route - Company A

Figure 5.5b Frequency distribution - Number of work assignments per pilot per period on the Inner Route - Company B

5.4 Number of Work Assignments per Pilot - Hydrographers Passage

Hydrographers Passage is a region of compulsory pilotage which extends east from Mackay and involves 10 – 14 hours of pilotage duties with the pilot required on the bridge for most of this time. For each of the companies across the three time periods, the mean number of work assignments per pilot are shown in Figure 5.6. Immediately noticeable from the figure is the small number of work assignments performed by Company C during the first time period. This result reflects the establishment of this company during this period.

Table 5.3 shows the between company and time period differences in the number of work assignments performed on Hydrographers passage per pilot. Inter-company differences related to the significantly fewer work assignments performed per pilot on this route by Company A (4.7) compared with Companies B (7.08) and C (7.44). A clear distinction of whether differences between Companies B and C existed was not shown by the post hoc analysis, possibly due to the sample size and variance associated with Company C’s data. Given that Hydrographers Passage is the only operational region for Company C, it was expected that pilots from this company would perform a substantial proportion of the work on this route.

Examination of changes across time indicated that there were no significant differences in the number of work assignments performed on Hydrographers passage across the 18 months (Table 5.3). This finding applied to each of the companies.

Figure 5.6 Mean number of work assignments per pilot on the Hydrographers Passage by Company and period (1 Jan 1996 - 30 June 1997)

Note: Data analysis based on n values shown in Methods (Section 3.2).

Table 5.3 Analysis of the mean number of work assignments per pilot on the Hydrographers Passage (1 Jan 1996 - 30 June 1997) (1)

1. Results of full two-way analysis of variance (ANOVA) model including company and period effects. p < 0.01 considered statistically significant.
2. Results of Tukey’s Studentised Range Test for post-hoc differences (Type I Experimental Error Rate = .01) n/a Post hoc testing not performed when main effects not significant.

Company (Main Effect). This term measures whether or not there is a difference between companies averaged across the three time periods.

Period (Main Effect). This term measures whether or not there is a difference between the periods averaged across the companies.

Company * Period Interaction. This term measures whether or not the difference between the companies is the same in each period. Equally, it measures if the difference between periods is the same for each company. When the interaction is statistically significant there is little interest in the main effects. In this case averages over one of the factors ignore important variations.

Figure 5.7a Frequency distribution - Number of work assignments per pilot per period on the Hydrographers Passage - Company A

Figure 5.7b Frequency distribution - Number of work assignments per pilot per period on the Hydrographers Passage - Company B

Figure 5.7c Frequency distribution - Number of work assignments per pilot per period on the Hydrographers Passage - Company C

5.5 Number of Work Assignments per Pilot - Great North East Channel

The Great North East Channel is a shipping route extending north east from Cape York. It is not an area of compulsory pilotage however, the Australian Maritime Safety Authority and the International Maritime Organisation recommend the services of local marine pilots be used when transiting this region (Australian Reef Pilots Ltd. 1996). Company C does not operate on this shipping route and therefore, was not included in the following analysis.

Figure 5.8 displays the mean number of assignments per pilot performed on the Great North East Channel, for Companies A and B across the time periods. Table 5.4 summarises the analysis of this data in terms of company and period differences. No significant differences between companies or over the three 6 month periods existed. This finding suggests that work on this shipping route was fairly evenly distributed between Companies A and B and did not change substantially over the 18 month analysis period.

Figure 5.8 Mean number of work assignments per pilot on the Great North East Channel by company and period (1 Jan 1996 - 30 June 1997)

Note: Data analysis based on n values shown in Methods (Section 3.2).

Table 5.4 Analysis of the mean number of work assignments per pilot on the Great North East Channel (1 Jan 1996 - 30 June 1997) (1)

1. Results of full two-way analysis of variance (ANOVA) model including company and period effects. p < 0.01 considered statistically significant.
2. Results of Tukey’s Studentised Range Test for post-hoc differences (Type I Experimental Error Rate = .01) n/a Post hoc testing not performed when main effects not significant.

Company (Main Effect). This term measures whether or not there is a difference between companies averaged across the three time periods.

Period (Main Effect). This term measures whether or not there is a difference between the periods averaged across the companies.

Company * Period Interaction. This term measures whether or not the difference between the companies is the same in each period. Equally, it measures if the difference between periods is the same for each company. When the interaction is statistically significant there is little interest in the main effects. In this case averages over one of the factors ignore important variations.

5.6 Duration of Work Assignments

Based on the work schedule files, the duration of work assignments, that is, the time on board ship, extended from the time of embarkation to disembarkation. However, following consultation with the Pilot Advisory Group members, it was determined that travel to and from the ship should be classified as work. Hence, as part of the analysis, work assignment duration was adjusted for travel. The results revealed a significant travel effect, indicating the inclusion of travel significantly (p < 0.0001) increased the duration of work assignments.

Figure 5.9 shows the duration of work assignments (ship time) in the three 6-month periods for each company separately. Table 5.5 summarises the analysis of assignment duration in terms of company and period differences. Significant company differences in work assignment duration were evident, with pilots from Company A working significantly longer assignments (45.6 hours) than pilots from Company B (40.8 hours) and Company C (13.9 hours).

While the substantially shorter duration of work assignments performed by Company C was expected due to the short transit time through Hydrographers Passage, the significant differences between Companies A and B is more perplexing. Both these companies operate in the same regions and therefore, it was anticipated that work assignments would be of similar duration. However, it is possible that the longer assignments of Company A may relate to personnel embarking on ships some days before entering the compulsory pilotage region (e.g. Brisbane) and then travelling on the ship without officially performing pilotage duties until the compulsory region. This practice may carry its own risk for fatigue potential, as the alternative living conditions experienced on board may not always be conducive to optimising sleep or alertness. Additionally, during busy periods this practice may not represent the most efficient use of a limited workforce.

When the average duration of work assignments of Great Barrier Reef pilots is compared to other groups of marine pilots, some of the unique features of pilotage in the Great Barrier Reef - Torres Strait region are highlighted. Long haul pilotage trips performed by pilots from the United Kingdom involve approximately 5.5 hours of navigating vessels (Shipley, 1978), while a typical pilotage assignment undertaken by Port Phillip sea pilots involves around 10 to 12 hours of pilotage work (Berger, 1984). Hence, it is apparent that the work assignments of Great Barrier Reef pilots are longer in duration than for either of the two comparison groups. This is especially true for pilots working for Companies A and B. The differences in the size of the area serviced by the pilot groups would seem to be one of the principle reasons attributable to this finding. Nevertheless, it is important to consider the possible effects of prolonged duty times on pilots. It could be anticipated that Great Barrier Reef pilots may be susceptible to greater levels of fatigue, as a direct consequence of the longer work assignments. It should be noted however, that work assignment times presented here only indicate the time between embarkation and disembarkation and do not differentiate between time working on the bridge and time spent resting. On tour log books will specifically detail bridge time and sleep periods per work assignments.

By presenting the duration of work assignments per six month period as frequency distributions (Figure 5.10a, and Figure 5.10b), some insight into the average figures reported above (Table 5.5) is obtained. Figure 5.10a clearly shows that several work assignments undertaken by Company A extended beyond a one week duration. In contrast, no assignments performed by Company B fell into this category (Figure 5.10b). This trend within the data most likely contributes to the longer average assignment duration reported for Company A. Although not shown, the short transit time through Hydrographers Passage resulted in all work assignments for Company C being less than 24 hours.

With regards to work assignment duration across time, no significant differences were observed. This temporal pattern was stable for all of the pilot companies (Table 5.5).

Figure 5.9 Mean duration (hours) of work assignments (ship time) by Company and period (1 Jan 1996 - 30 June 1997)

Note: Data analysis based on n values shown in Methods (Section 3.2).

Table 5.5 Analysis of the mean duration (hours) of work assignments (ship time) (1 Jan 1996 - 30 June 1997) (1)

1. (1) Results of full two-way analysis of variance (ANOVA) model including company and period effects. p < 0.01 considered statistically significant.
2. (2) Results of Tukey’s Studentised Range Test for post-hoc differences (Type I Experimental Error Rate = .01) n/a Post hoc testing not performed when no significant man effects.

Company (Main Effect). This term measures whether or not there is a difference between companies averaged across the three time periods.

Period (Main Effect). This term measures whether or not there is a difference between the periods averaged across the companies.

Company * Period Interaction. This term measures whether or not the difference between the companies is the same in each period. Equally, it measures if the difference between periods is the same for each company. When the interaction is statistically significant there is little interest in the main effects. In this case averages over one of the factors ignore important variations.

Figure 5.10a Frequency distribution - Duration (hours) of work assignments (ship time) per period - Company A

Figure 5.10b Frequency distribution - Duration (hours) of work assignments (ship time) per period - Company B

5.7 Duration of Breaks between Work Assignments

Breaks between work assignments potentially represent a period of time during which pilots can recuperate from a work assignment and prepare themselves for their next voyage. For pilots who reside in the operational region (Northern and Central Queensland), breaks may be spent at home. However, in many situations breaks are spent in alternative accommodation such as hotels, motels and/or pilot houses.

Within the present work schedule files, breaks between work assignments were determined from the date and time of disembarkation to the date and time of the next embarkation (non-ship time). Travel to and from ships significantly decreased (p < 0.0001) the duration of breaks between assignments and hence, breaks were travel adjusted. In some circumstances, travel between ports may also be undertaken during assignment breaks, which in turn would further reduce break duration.

In an attempt to ensure that pilots receive adequate time to recover between work assignments, the Australian Maritime Safety Authority have established regulations stipulating minimum rest breaks between consecutive pilotage assignments performed in the Great Barrier Reef - Torres Strait region. These regulations state that prior to any work assignment, a minium of 12 consecutive hours of rest, excluding travel must be taken by the pilot, except for when the Inner Route passage is to be piloted, in which case at least 24 consecutive hours of rest, excluding travel is required (AMSA, 1997).

Figure 5.11 illustrates the average duration of assignment breaks in the three time periods for each company separately. A summary of the analysis of this data in terms of company and period differences is shown in Table 5.6. There were significant differences between the companies, with pilots from Company A recording breaks of 52.8 hours, pilots from Company B recording breaks of 48.9 hours and pilots from Company C recording breaks of 72.0 hours. These results seem to suggest that all of the Companies are complying with the minimum break duration regulations stipulated by the Australian Maritime Safety Authority.

Mean data however, fails to highlight the variability which exists within any data set and hence, to assess variability, the duration of breaks between work assignments have been presented as frequency distributions (Figures 5.12a, Figures 5.12b and Figure 5.12c). From these figures, it can be identified that in the majority of situations, rest breaks do conform with the Australian Maritime Safety Authority regulations. Most work assignments are interspersed by at least 24 hours of non-work time, thereby more than adequately meeting the 12 and 24 hour minimum time periods state in the regulations. However, it is also evident on the distributions of Companies A and B that there were a number of times when breaks between work assignments failed to comply with regulations. While extenuating circumstances may have existed, that less than 12 hours rest between work periods was obtained on some occasions is somewhat disconcerting.

A recent investigation revealed that during work assignments, both quantity and quality of sleep achieved by Great Barrier Reef pilots tends to be compromised. For instance, thirty-one percent of Great Barrier Reef pilots attain an average of less than 4 hours sleep per 24 hours while working, while 55 percent rated their sleep as fair, poor or very poor in quality (Parker et al. 1997). When these findings are viewed in combination with the results of studies investigating the effects of restricted sleep on mood and performance, a highly undesirable picture emerges. Restricted sleep has been shown to cause significant reductions in cognitive, vigilance and memory performance and increased levels of fatigue (Dinges et al. 1997; Tilley & Wilkinson 1984; TSB 1997; Wittersheim et al. 1992). Additionally, it seems that at least 2 nights of recovery sleep are required following a period of sleep restriction, before complete recovery is achieved (Dinges et al. 1997; Morris & Miller 1996; TSB 1997; Wittersheim et al. 1992). These findings place doubt over the current regulations stipulating minimum rest breaks of 12 and 24 hours (AMSA 1997) as even with 24 hours rest between work assignments, the opportunity to achieve 2 nights of recovery sleep is not available. Furthermore, that in certain situations Great Barrier Reef pilots are only achieving 12 or less hours recovery time between work assignments suggests there may be a heightened risk of performance impairment and/or fatigue development. Accordingly, a review of the current regulations and future monitoring of rest break duration may be warranted.

At no time during the 18 month analysis period did the rest breaks experienced by Company C fail to meet current regulations (Figure 5.12c).

By considering average assignment and break duration together (Table 5.5 and Table 5.6), an approximation of the work (ship time) to rest (non-ship time) ratio for each of the pilot companies was calculated. Both Companies A and B recorded work to rest ratios of around 1 : 1.2, whereas Company C recorded a ratio of 1 : 5.2. Thus, relative to work assignment duration Company C personnel experience much longer periods of time off between assignments compared with either of the other two pilot groups.

That both Companies A and B recorded the same work to rest ratio highlights the similarity between the two groups in terms of pilotage operations. Based on results so far, it would appear that Company B personnel perform more assignments of shorter duration with shorter rest breaks, whereas Company A pilots perform fewer but longer work assignments with longer rest breaks. A point of interest may be whether a greater number of shorter assignments with shorter breaks is more fatiguing than fewer longer assignments with longer breaks. That is, whether more frequent travel associated with a greater number of shorter assignments is more fatiguing than exposure to the on-ship environment for longer periods.

Comparisons of the work to rest ratios of Great Barrier Reef pilots with the work to rest ratios of other groups of marine pilots reveal some interesting findings. For instance, the work to rest ratio calculated from data on long haul pilots from the United Kingdom (Shipley 1978) was approximately 1 : 9.25. Thus, the rest periods relative to work periods were more than six times longer for the pilots from the United Kingdom, compared with Great Barrier Reef pilots working for Companies A and B. These findings need to be interpreted cautiously as distinct differences between the two pilotage groups exist in terms of the size of the region serviced, work assignment duration and work roster systems. However, the much shorter period of rest experienced by the majority of Great Barrier Reef pilots seems to support the previous suggestion that Great Barrier Reef pilots may have an increased susceptibility to experiencing fatigue as a consequence of their work conditions, as compared to some other groups of marine pilots.

With regards to changes across time it was evident that the duration of breaks between work assignments was significantly shorter during the second time period as compared to the first and third periods. This finding is consistent with the increased work availability reported during this middle 6 month period.

Figure 5.11 Mean duration of breaks (hours) between work assignments (non-ship time) by Company and period (1 Jan 1996 - 30 June 1997)

Note: Data analysis based on n values shown in Methods (Section 3.2).

Table 5.6 Analysis of the duration of breaks (hours) between work assignments (non-ship time) (1 Jan 1996 - 30 June 1997) (1)

1. Results of full two-way analysis of variance (ANOVA) model including company and period effects. p < 0.01 considered statistically significant.
2. Results of Tukey’s Studentised Range Test for post-hoc differences (Type I Experimental Error Rate = .01).

Company (Main Effect). This term measures whether or not there is a difference between companies averaged across the three time periods.

Period (Main Effect). This term measures whether or not there is a difference between the periods averaged across the companies.

Company * Period Interaction. This term measures whether or not the difference between the companies is the same in each period. Equally, it measures if the difference between periods is the same for each company. When the interaction is statistically significant there is little interest in the main effects. In this case averages over one of the factors ignore important variations.

Figure 5.12a Frequency distribution - Duration of breaks between work assignments (non-ship time) per period - Company A

Figure 5.12b Frequency distribution - Duration of breaks between work assignments (non-ship time) per period - Company B

Figure 5.12c Frequency distribution - Duration of breaks between work assignments (non-ship time) per period - Company C

Accessibility | Copyright & Privacy
ABN 65 377 938 320