Try these
great search terms!
Form Design
Data capture
Survey Services
Data Logging
4. Quality of Data
Gathering raw figures and data is all very well - but how do you know that the data is not rubbish?
You have to think about the quality of the data - how can things go wrong?
Consider these situations:
|
Unreliable questionnaires
If an inappropriate individual has been asked the questions, then the data is most likely to be unreliable. For example asking a five year old their views on the latest interest rate rise is not likely to be valid! Unless the questions are carefully worded, then they may be misunderstood by the public and so the data once again becomes unreliable. Avoid ambiguous questions.
|
|
|
Incomplete data
Goods can leave a store by many different ways - the main one is by sales which are recorded by bar code readers at the till. So if the manager made this simple calculation based on sales data :- |
|
Stock left = Original stock - Stock sold
then is she correct?
No - the data is incomplete because there are other ways of reducing stock.
 |
Theft - stock can disappear! |
 |
Spoilage - Perishable goods have to be thrown away after their sell-by date. |
So unless that manager gathers data on these methods as well, her stock estimate would be incorrect.
GIGO
Garbage In, Garbage Out : Meaning that poor data coming in, means poor information coming out.
| If you are gathering data by means of instruments then calibration against a standard is essential before you start taking measurements. |
|
|
|
If you are gathering data using an on-line form, then there has to be error checks in place to make sure that sensible data is being provided. What would you do with a customer who tells you they are called 'Mickey Mouse'?
|
| If you are loading data into a database, then the system should check for obvious errors as the data is read in i.e. data validation. |
|
Copyright © 2006 www.teach-ict.com