Mobile Edition
Print
Get the Mag
Weekly UpdatesThe data were analysed with the SPSS software using the percentile method, Kendall's coefficient of concordance, the Mann-Whitney test and severity index analysis. The formula for the severity index is given as follows by Elhag and Boussabaine (1999):
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII],
where: S.I. is the severity index; [f.sub.i] is the frequency of response; [w.sub.i] is the weight for each rating (i.e. rating in scale/number of points in a scale), and n is the total number of responses. The value ([f.sub.i] x 100)/n is the valid percentage as computed by SPSS.
5. ANALYSIS AND DISCUSSION OF RESULTS
Of the 655 questionnaires distributed, 406 were returned. This represents a response rate of 61.98%, which is very good, according to Ellhag and Boussabaine (1999) and Idrus and Newman (2002) who have expressed the view that a response rate of 30% is good enough in construction. In addition, the 406 responses obtained gave the sample a confidence interval better than [+ or -] 5.0 (De Vaus, 1996). The tenants' mean length of stay in the houses was 8.44 years, which indicated that on the average the tenants had lived in the house long enough to provide detailed information on the maintenance history of their houses. The results are presented on tenants' maintenance awareness, responsibility, priority preferences, performance rating of and satisfaction with the state of maintenance of their houses.
5.1. Tenants' Maintenance Awareness and Responsibility
As major stakeholders in housing maintenance, tenants are expected to understand and appreciate the need for maintenance. In Table 1, only 8 respondents (2%) had no idea why maintenance was necessary, while most of them (77.8%) believed that maintenance was necessary to keep the house safe for habitation. This high level of maintenance awareness may encourage tenants to respond promptly to the detection of defects in their houses. In fact, Table 2 shows that more than a quarter (28.6%) report faults promptly while about half (44.5%) even fix minor faults themselves.
The high level of maintenance awareness and responsibility demonstrated by the tenants is reflected in the fact that 50.7% of the respondents believed that it was their duty to use the houses with care while only a negligible number 1.7% felt it was none of their business to contribute. Some tenants were even prepared to do some maintenance work themselves (44.6%) or provide materials when the maintenance department did not have them (30.8%).
5.2. Tenants' Priority Preferences
All over the world, the dwindling resources available for maintenance (including Housing maintenance) in the face of ever-increasing maintenance demands requires that maintenance needs be prioritised to achieve the best value for money (Ramamurthy, 1990; Berger et al., 1991; Shen et al., 1998; Vijverberg, 2000). Traditionally, maintenance prioritisation has been done by maintenance departments without considering the views of tenants. Yip (2001) has argued in favour of tenant participation in this vital exercise to enhance tenants' satisfaction with maintenance systems. Towing this line, tenants were asked in this study to rank 16 common building defects in order of priority. In Table 3, the significance value of Kendall's coefficient of concordance is 0.000 (i.e. < 0.05), indicating that there was agreement (at 5% significance level) among the tenants in their priority preferences. These results enable us to reject the null hypothesis that "There is no agreement among tenants in their maintenance priority preferences" and accept the alternative that "There is agreement among tenants in their maintenance priority preferences".
Tenants' priority preferences were then compared with those of the maintenance departments in Table 4 to see if there was any harmony between the two. It is not surprising that both the tenants and the maintenance staff ranked roof structure number 1, as it is apparent to both groups that a collapsed roof exposes other parts of the building to the elements and can endanger both lives and property. On the other hand, both wall tile failure and floor tile failure, which pose no such dangers as for collapsed roof structures, were ranked number 16 (last) by the tenants and the maintenance departments respectively. These rankings by the tenants also show that they have a high level of awareness and understanding of the rationale for maintenance.
To determine whether the rankings of tenants and the maintenance departments differ significantly, the nonparametric method of the Mann-Whitney test is applied in Table 5. It is used to test the null hypothesis that two populations have identical distribution functions against the alternative hypothesis that the two distribution functions differ only with respect to location (median), if at all. The results in Table 5 indicate that only in three defects (defect in roof structure, floor tile failure and damaged taps/stop valves) out of the sixteen defects are there significant differences between the priority preferences of tenants and the maintenance departments (p < 0.05 at the 5% significance level). Therefore we accept the null hypothesis that "There is no significant difference between the maintenance priority preferences of tenants and the maintenance department" and reject the alternative hypothesis that "There are significant differences between the maintenance priority preferences of tenants and the maintenance department".
The fact that there is agreement among the tenants in their priority ranking of building defects, and there is no significant difference between the tenants and the maintenance departments in their priority preferences augurs well for the prioritisation of maintenance works in two ways. First, agreement between tenants on the one hand and the maintenance departments on the other ensures that the two sides do not have conflicting expectations. Secondly, agreement among tenants themselves makes it possible for maintenance management to fix priorities acceptable to the generality of the tenants.
5.3. Maintenance Performance Rating and Tenants' Satisfaction
As stated earlier, the level of tenants' satisfaction is an index of maintenance performance. For tenants' satisfaction to be used as a measure of housing maintenance performance, it is important that tenants be in agreement in their assessments of the various attributes of maintenance satisfaction. Also, for tenant ratings of their satisfaction to be objective it must be based on the actual state of maintenance of their dwellings and not on extraneous factors. To examine these issues, two null hypotheses were postulated as follows:
* There is no agreement among tenants in their satisfaction rating of housing maintenance management.
* There is no significant correlation between users' perception of the level of maintenance of a particular building and the prevalent level of user satisfaction.
The first hypothesis is tested using Kendall's test of concordance (Table 6). For the significance level p < 0.05, we reject the null hypothesis that "There is no agreement among tenants in their satisfaction rating of housing maintenance management" and accept the alternative hypothesis that "there is agreement among tenants in their satisfaction rating of housing maintenance management".
With the results above we proceed in Table 7 to analyse tenants' satisfaction with the various attributes of the housing maintenance systems, using a scale from 1 = Very dissatisfied to 7 = Very satisfied. The results show that the quality of the environment and surroundings is the maintenance attribute most satisfactory to the tenants while the level of maintenance backlog is the least satisfactory. The high, unsatisfactory level of maintenance backlog stemmed from the fact that in the past five years the maintenance departments received just about 15% of their actual budgetary requirements from the institutions.
Table 8 shows that out of the 391 respondents, 129 (33.1%) rated their satisfaction with the overall maintenance of their dwellings below average while 111 (28.4%) rated it above average. Only 14 respondents (3.6%) were "Very satisfied" while 37 (9.5%) were "Very unsatisfied". The majority of 151 (38.6%) were only averagely satisfied.
Applying the formula for severity index (explained in the methodology) to the results in Table 8, an overall tenant satisfaction index (TSI) of 0.55 is obtained, which on a scale of 0 to 1 is just average.
In Table 9, the overall rating (from 1 = Very bad to 7 = Very good) of the state of maintenance of the houses by the tenants is presented. Again, using the formula for severity index earlier given, a state of maintenance index (SMI) of 0.57 is calculated from the results in Table 9. The SMI is just above average on a scale of 0 to 1.
The second null hypothesis that "There is no significant correlation between users' perception of the level of maintenance of a particular dwelling and the prevalent level of user satisfaction" is tested in Table 10 by running Pearson's correlation between the tenants' overall rating of the state of maintenance of their dwellings (from very bad to very good) and their satisfaction with the overall maintenance of their dwellings.
The analysis in Table 10 indicates a positive correlation between the two variables. For p < 0.01, the correlation is highly significant at the 1% level. Hence we reject the null hypothesis that "There is no significant correlation between users' perception of the level of maintenance of a particular dwelling and the prevalent level of user satisfaction" and accept the alternative hypothesis that "There is significant correlation between users' perception of the level of maintenance of a particular dwelling and the prevalent level of user satisfaction". This implies that the tenant satisfaction index of 0.55 obtained represents the true level of tenant satisfaction based on the actual state of maintenance of the dwellings and not on extraneous factors. The maintenance departments can therefore rely on the results as feedback from the tenants to guide their decisions to improve performance
FEED