Reliability
by Sam Cook
In scientific research the data collected is seen as being reliable if other researchers using the same methods to test the same material get the same result – my boiling point of water experiment explains – so reliability refers to 
replication. If the results are consistent the researcher can draw generalisations.
The best way to understand this is with schools. If school A and school B both enter students for the same GCSE Maths examination and a sociologists examines the examination results (the data) in order to find out which school is the best at Maths the sociologist could say the school with the lowest Math passes is the lesser school.
Such a statement is known as a generalisation – those schools with lower GCSE Maths grades are poorer performing schools than those with higher GCSE Maths grades. A generalisation can be made when the data gathered is seen as valid, in this example it is produced by an exam board.
This process of gathering statistical data from exam scores is seen reliable because the sociologist can repeat the process next year and through successive years (a point worth noting here is a research method can be reliable without it necessarily being valid).
Research producing quantitative data generates more reliable data than qualitative methods this is because an unstructured interview for example is by its very nature hard to replicate. Nevertheless the data gathered from an unstructured interview is seen to be high in validity (according to interpretivists).
sociologytwynham.com 
Leave a comment