Then , the personalisation strategies module assessed any modifications to the dwellings including size ( surface area ) and materials for extensions ( wood , masonry or concrete ), finishing works ( paint , paint with primer , ceramic tiling , paper , vinyl , carpet , wood tilling , and wood panels ), date of construction , builder ( self-construction , informal or formal contractor ) and final use . The interviewees were also asked to identify the modification they considered to be the most important and the motivations behind it through open questions , whilst in parallel a photographic survey was used to collect detailed information on qualitative aspects of these modifications . The information was later coded and only permanent to semi-permanent modifications were included in the analyses . The extent to which each dwelling was modified was classified as ‘ scarcely modified ’ when only cosmetic modifications were found , ‘ averagely modified ’ when they included anchored furniture or minor changes to internal partitions , and ‘ highly modified ’ when they included extensions or significant changes to the layouts . The workmanship quality of these modifications was coded as ‘ high-end ’ when the results were close to professional , ‘ average ’ when the results were sub-standard and ‘ low-end ’ when they evidenced significant qualitative problems or use of scarp materials .
The dwellings were randomly selected before the fieldwork , and the sample size was determined for a confidence of 99 % and an interval of 15 % for each housing complex for a total of 130 cases ( Table 1 ). The survey was conducted on site between the months of August and September 2014 , and only heads of household and / or their partners were approached and asked to voluntarily answer a questionnaire whose results were anonymised before the analyses .
After descriptive statistics and preliminary analyses through visual methods , Spearman ’ s Rank Order Correlation was used to identify statistically significant relationships between ranked variables where rs = 0.00 to 0.19 was considered as negligible , rs = 0.20 to 0.29 as weak , rs = 0.30 to 0.39 as moderate , rs = 0.40 to 0.59 as strong , and rs = 0.60 to 1.00 as very strong correlations . Then , Kruskal-Wallis H test was used to determine the statistical significance of observed differences in satisfaction and personalisation strategies among groups ( i . e . dwelling types , developers , and general household characteristics ). All the analyses were conducted using IBM SPSS Statistics v22 .
Table 1 . General characteristics of the sample population by case study housing complex |
|
case 1 |
case 2 |
case 3 |
case 4 |
developer |
non-profit |
non-profit |
for-profit |
for-profit |
date of construction |
2011 |
2012 |
2013 |
2013 |
dwelling type |
house |
Apartment |
house |
apartment |
surface area |
58 m2 |
57 m2 |
45 m2 |
55 m2 |
number of bedrooms |
3 |
3 |
2 |
3 |
number of units |
32 |
64 |
46 |
104 |
surveyed cases |
23 |
35 |
29 |
43 |
Results
Demographic Characteristics The mean household size was close to the national average with 3.79 users per dwelling , whilst single-resident units comprised only 1.4 % of the total and 2.1 % of the dwellings were unoccupied at the moment of the survey . 21.51 % of the total households had more than one family nucleus , and the mean number of workers per household was 1.27 , with 56.5 % of the total having a single worker per dwelling , 29.6 % two workers , 3.7 % three workers , and 10.2 % no active worker at
492 ZEMCH 2015 | International Conference | Bari - Lecce , Italy