Reliability

Quality of data one gets depends on a number of factors. How do you know that a sample of 2 from 2000 over students that you have taught represent a 'true' picture of one behavior. Obviously, 2/2000 is only 0.1% of the population. So, the results surely cannot be conclusive. This is an aspect called sampling and it is especailly important in research, and probably in life as well. So, the next time someone tells you that 'that person' is not up to 'standards' please ask yourself. Is the numberof samples representative, sufficient, valid and reliable. Reliability can be checked thorugh content validity, construct and criterion validity. An one more thing try not to generalise from unfounded figures and 'facts'.