Preliminary Data Analysis

We collected 972 questionnaires but two of them had little more than the first page completed and were discarded leaving 970 for analysis.

As might be expected, there were gaps with some questions unanswered and some responses could not be used when instructions were not followed. Most notable in this aspect was question 2.

2	Please enter the number of people in your household in each of the following age groups	0 to 10		11 to 16
2		17 to 60		over 60

For this question we had to discount 51 responses.

Some other questions had errors in responses, often this occurred when the question had asked for an order of priority, e.g. answer using 1, 2 or 3 and ticks were simply placed in the boxes. Occasionally for such questions as ‘list the three most important factors’ more than three choices were made.

Those answers which did not follow the rubric were dealt with in two ways; where the number of inappropriate answers was a very small proportion of the total they were simply ignored but in those questions where the proportion was higher, perhaps because the question was badly worded, the number that could not be analysed was included in the figures.

Generally, the response rate for individual questions was comfortingly quite high and as one might expect the rate was at its highest where the question was deemed to be of more importance to the individuals responding.

The easiest response rates to measure were those questions which demanded a simple yes/no/no opinion answer. Of 27 of those looked at, 12 had over 95% response with the highest at 99% for question 5a asking about the volume of traffic in Willingham. A further eight had response between 90-95%. The unusually low response rates were between 60-65% and occurred with the questions about the Guided Bus, perhaps because people had not yet considered whether they would use it or how they might travel to the station at Longstanton and whether the route would suit their needs.

The analysis was undertaken in three parts. In the first phase those questions which involved filling boxes, whether they be yes/no/no opinion or put in order or list which applies, were entered on an Excel spreadsheet and our thanks go to the many volunteers who undertook the task of entering data. The volunteers were people living in the village and they sent their individual entries to be compiled into one master list. The analysis of responses was made fairly easy using Excel.

Those questions which asked for open comment were analysed in the second and third phases. The open responses were divided into two types. In the first type covering those open response questions we thought every single comment should be looked at carefully so as to be able to gain both strength of feeling and to trap the many good ideas and suggestion which came forth.

In the final group we put questions such that we had already formed a good idea of what might be said. A typical example would be asking where speeding traffic posed a problem. This small group of questions were sampled from the 970 questionnaires. The sample was chosen by postcode and in proportion. Over 25% were analysed in this final batch.

Below is a shortened version the first part of the analysis which covered the simple response questions. The remaining two parts which contained all the comments will be reported on later. The comments had to be sorted and categorised to make them more readable as they ran to very many pages of script.

This preliminary analysis is also available in PDF format at the link below.

Preliminary Analysis Report [PDF - 302kb]