Reasons for screening data

•     Garbage in-garbage out
•     Missing data

    
    a. Amount of missing data is less crucial than the pattern of it.

• If randomly scattered not a problem/ nonrandom patterns limit the use of data (can't generalize results)

• Extreme values or outliers - cases with extreme values on one or a combination of variables that can potentially distort the results of the analysis. Ascertain that the data fulfills the basic assumptions for statistical techniques:  

a. Data being normally distributed
    b. Linear relationship between variables Homoscedasticity