Not all those interviewed in the Intangibles Survey were able or willing to complete the GHQ12 questionnaires. In the flooded interview, the request to provide these responses came more than half way through the interview after the respondents had answered questions on flooding and its impacts but before the willingness to pay questions. Among those flooded, a total of 814 answered the current GHQ12, giving a non-response rate of 17.4%. Most of those flooded respondents who provided answers to the current GHQ12 went on to give information about their worst time. Worst time GHQ12, responses were obtained from 810, giving a non-response rate of 17.8%. In the at risk sample, the respondents were faced with a shorter questionnaire and response on the current GHQ12 was higher in this sample, with responses obtained from 485 at risk respondents, giving a non-response rate of 8%.
Response was much higher on the questions rating the stress of the flood event and the overall severity of the effects of the flood. The number of flooded respondents providing answers to these questions was 972 with only 11 non respondents for stress and 973 with 10 non-respondents for overall severity.
Thus, it should be noted that the responses to the variables are drawn from different groups of respondents.
Correlations between vulnerability variables
The following table (Table 6.3) shows the correlations between current and worst GHQ12 scores (both GHQ and Likert scoring) and the other measures of vulnerability to give an indication of the extent to which the different measures are measuring the same thing. Not surprisingly, the two methods of scoring the GHQ12 produce scores that were very closely correlated.
When the current and worst GHQ12 scores were correlated with the subjective severity measures of stress and overall severity, the worst time scores were more closely related to stress and the overall severity than the current GHQ12 scores. This is true whatever method of scoring the GHQ12 is adopted. The worst time, stress and overall severity vulnerability measures were all placed closely in the context of the recent or worst flood event, whereas the current GHQ12 questions administered without reference to the flood event, were not. However, the current GHQ12 scores were moderately strongly correlated with the worst time scores under both methods indicating a link between current and worst time health.
Stage_at_which_the_health_and_other_impacts_of_flooding_were_worst'>Stage at which the health and other impacts of flooding were worst
The responses to the question “At what stage during or after the flooding were the health impacts the most severe or worst for you personally? Please think about health in the broadest sense to include physical, mental and social well-being” provide a guide as to when flooded respondents considered themselves to be at their most vulnerable. Figure 6.2 shows the stage when the health effects were most severe. A significant proportion (76 or 8%) did not name a stage. There was a wide range of responses on the time when the impacts were worst. For half the respondents who did respond, the worst stage occurred early on, during or within a month of the event in the immediate aftermath as they came to terms with what had happened. However, there were others for whom the worst effects were not felt until months after the flood itself during the recovery period. Respondents were also asked about the duration of the worst stage. A third of respondents did not provide information on this. Of those that did, a majority reported the worst stage lasted no more than two or three weeks. However, there were others who reported suffering severe effects for months.
Figure 6.2: Percentage reporting stage at which the health impacts were most severe: Intangibles Survey
Men and women showed the same pattern of response on the stage at which the impacts were worst for them. There were also no significant differences in the responses according to the social grade of the respondent. However, there were significant differences in the responses with age (Chi-square; p<0.01). In particular, there was a trend for the proportion citing ‘during the flood event itself’ to increase with age and this was the most common response of the 75 and over age group, given by 25% of them. People aged under 55 most commonly reported the first week or two as having most impact.
The stage at which the impacts were worst for flooded respondents also varied according to the extent (Chi-square; p<0.01) and depth (Chi-square; p<0.05) of main room flooding. For those minimally affected i.e. with no main room flooding, or only one main room affected and those flooded to a low level (<10cms), the worst period occurred early for most people. Their most common response was ‘during the flood event itself followed by the first week or two after the flood. There were few of the minimally affected for whom impacts peaked later. Those worst affected i.e. with four main rooms flooded or a metre or more of flood waters in the home commonly cited one to three months after the flood as well as the first week or two as their worst time and their range of responses was quite varied. For those whose extent and depth of flooding fell in between, the first week or two was a common response but again the peak time occurred at varying times after the event. Table 6.4 shows the mean scores on the vulnerability variables according to the reported worst stage.
Table 6.4: Mean scores on vulnerability variables according to the reported worst stage: Intangibles Survey