Accuracy and Reliability

Once you have stated and explained your conclusions, you need to start thinking about the "bigger picture" - where does your investigation fit with the wider field of Geography?  What is the meaning of your research and what is the worth of your results?


This term refers to your data - and particularly to the precision with which it was collected.  What were the likely sources of error? Is your data specific and precise or are there ways in which there might be mistakes in the data collection, or problems with the data? 

You will need to comment on how accurate your results are, and the likely problems which have limited this.


This is more to do with your conclusions - are they believeable? If you repeated the investigation would you get the same results?  Can your conclusions be accepted as an overall trend or theory?  To quote an old advert (and more recent comedy series), is 8 out of ten cats enough?

The likelihood is that your investigation is to an extent accurate and reliable, but that there were problems with your sampling and data collection that make your conclusions less than valid.  Here are some questions to ask yourself (not all will be relevant to your study)...

  1. Is one transect enough?
  2. Are 20 sites enough?
  3. Was one day enough?
  4. Are 30 questionnaire results sufficient?
  5. How reliable are peoples' opinions?
  6. What is the margin for error when estimating % cover in a quadrat?
  7. Was one survey location appropriate?
  8. Did you identify every different grass species correctly?
  9. Was your sampling strategy truly systematic?
  10. Is the temperature sensor calibrated correctly? Did you check?

If the answer to any (or all) of these questions is "no" (hint: it should be!), then you have ideas as to how to go about commenting on the accuracy and reliability of your data.



For each aim, identify at least one way that the accuracy of your data and the reliability of your conclusions can be questionned. 

Try to identify ways that these problems can be overcome.