Evaluation

Evaluation

This is the final section of your CA report... However, it is worth considerable marks (one sixth of the total available!), so make sure that you produce an effective evaluation by following the guidance below.

The table below describes the expectations for each level for the evaluation component of your report. 

NOTE: you have to achieve all the elements of the levels below before you can be awarded the higher levels.

Table showing elements for the Evaluation of CA. Note that Level 1 criteria are in column 2, level 3 in column 3 and level 3 in column 4.

In this section you need to DEMONSTRATE that you understand the link between the way in which data is collected (quality and appropriateness of the methodology and how accurately the techniques were applied) data collected (eg its accuracy) and the conclusions drawn. 

For example, if the data gathering METHODS meant that inaccurate DATA was collected, the results/CONCLUSIONS would probably be unreliable.

It is likely that, even with the best planning, there will be some weaknesses in the methods and, therefore, the data collected and so we must acknowledge that the conclusions have to be qualified to some extent. This does not mean that they are not valid, just that we have to be cautious in ascribing general applicability to our findings.

Typically, a good quality evaluation will consist of something along the lines of...

Evaluation.
As with all investigations in which primary data is collected, the quality and reliability of the conclusions conclusions depend on the appropriateness of the methods used to collect data, how well the methods were applied by the team of field data collectors and, thus, the accuracy of the data itself.

In this investigation, the methods used were fairly simple and should have provided data of appropriate accuracy for the purposes of identifying the presence or absence of a CBD in Ealing. However, because data was collected as a class with groups being responsible for different transects there may have been some inconsistency in the way in which data was collected. This, though, does not appear to have had a significant effect on the collated data and thus the results are likely to be reasonably accurate. However, one problem did occur with transect G. The team responsible for this transect failed to collect or to collate all the necessary data. This meant that, overall, less data was available and so conclusions based on this would be less reliable or representative than if the full set of data were available. It is acknowledged that the availability of data for this transect, might have made a difference to the conclusions that have been drawn as it constitutes 1/7th of all the data and this is statistically quite significant. The availability of the missing data would have made the final conclusions more accurate and probably more reliable.

In the above paragraph, the overall appropriateness of the METHOD is outlined, but a weakness in the actual DATA COLLECTION process is also described and identified as possibly undermining the final RESULTS.

This weakness might have been resolved by re-doing the field data collection and ensuring that the missing transect was completed too. However, this has not been possible due to the time constraints.

Here a possible improvement has been suggested. So far, the requirements for level 1 and have both been met. Furthermore, a specific problem has been identified (level 2) and the effect of the improvement has been suggested ie more accurate and reliable results (level 2).

Further issues regarding the validity of the results are related to the fact that this is one small study, conducted at a particular time and place. As such, the findings really only relate to the sample area for the day and time of the survey. It is possible that, for some data (such as pedestrian counts) Different results might be collected if the study was repeated at different times or days. 

Here the wider applicability of the findings is questioned with respect to the scope of the study/data collection period. This would meet level 3 requirements concerning the validity of the results. The same is achieved by the following paragraph...

Other data such as building height would not change over anything but the much longer timescale. Even so, if different transects were selected, this might have led to different patterns being detected. To some extent the chance of this happening has been mitigated by selecting a range of different streets to sample along. Consequently results should be more reliable and representative of the sample area.

The following paragraph demonstrates how broader geographical understanding can be shown within the evaluation... Models in geography are generalisations and while this makes them broadly applicable to a range of similar situations... they rarely are totally accurate for every situation in the real world. The most reliable models, though, are likely to have been developed and tested in a range of different places and at different times...

When this kind of research is carried out, it is often extrapolated to other similar situations. In this case, it might be suggested that the findings form Ealing could be applied  to other sub-centres of London (or even other metropolitan centres) that have a similar history and function. So, perhaps, Hammersmith, Brentford, Shepherds Bush etc also show signs of land-use zoning with a distinct CBD at their original centres and transition into residential areas. However, this must be done cautiously, if at all. If it was intended to apply the findings to other sub-centres, it would be necessary to carry out similar research in other places and at other times confirm the general patterns. The more places found that follow the patterns identified in Ealing, the more reliably the conclusions can be used as a general model.

Although there are evident weaknesses in this study as referenced above, and while it is recognised that further research at different times and at different locations would improve the reliability and applicability of the findings, the conclusions of this report are regarded as accurate for the time and place of the data collection. The problems described above are judged not to significantly undermine the conclusions although the findings are presented with a degree of qualification.