MR final exam review

Normal distribution, mean, variance

Add picture for N(M,variance)

Editing

• establishes policies regarding, incomplete answers, wring answers, lack of interest, etc.

severity of the problem

How to handle incomplete questionnaires depends on the what?

Coding

Technical procedures which raw data is transformed into symbols (usually and preferably numbers)

Steps to coding

1. Specify the categories

2. Assign code numbers to the categories

3. Prepare a codebook that contains general instructions indicating how each item of data was codes

Codebook

Hold all the codes and instructions for your data.

Variable transformation

• Variable re-specification

• scale transformation

• weighting

Variable re-specification

Collapsing into fewer variables

• data reduction

• Not recommended in the data entry stage

Scale transformation

For comparability among variables

• when you want to compare sales with. prices

• Standardization is the most common technique

Weighting

Assigning some amount of weight according to a prespecificed rule

• Do you have some more important respondents?

• Be cautious, weighting should be explicitly recorded

Plots

Used to help you understand your data

• identify data outliers and errors (data cleaning)

• To find basic relationships

• To provide a basis for evaluating more formal analysis

• Scatter plots, Box plots (min, q1, q2, q3, max)