Laserfiche WebLink
<br />17 <br /> <br />Since those interviewed were seleGted on wil"lingness to participate <br />rather than a more objective method, thl,re are certain biases 'in the <br />sample and, therefore, in the findings. The college newspaper ads yielded <br />response from young students rather than from another sector of the <br />population-at-large. <br />Learning the last action taken by a flood victim is a difficult <br />task. The researcher was careful not to increase stress to those famil ies <br />who suffered losses; therefore, no relatives of those who died were <br />interviewed. <br />The research effort has been compared to gathering and sort'ing of <br />clues in order to unravel a detective story. For example, a tourist <br />would say he saw the driver of a Winnebego camper with three people in <br />it panic and drive right into the f'lood and smash against a bridge. <br />Another person would say he saw the Winnebego and noticed there were two <br />women and one man in the camper; he thought the woman who was driving <br />was in her sixties. With this information and the list of the victims, <br />it was possible to tell who the victims were if none of the other victims <br />had similar characteristics. If they did, the search would continue. <br />Many assumptions were involved and they bias the results. <br />Knowing only imprecisely the population in the canyon the night <br />of the flood prohibited the choice of a fully random sample. In addition, <br />people who did not wish to share their experiences were not included in <br />the study. To compensate partially for this bias, the study was designed <br />to include all areas of the canyon and to include equally groups which <br />received warnings and groups which did not. <br />Analysis <br />The information gathered in the study was recorded on questionnaires <br />filled in during open-ended interviews or following conversations with <br />