Research Methods in MIS: Data Handling and Reduction

1
Research Methods in MISData Handling and
Reduction

• Dr. Deepak Khazanchi

Adapted from Sproull (1995).
2
Preparation for Data Handling

• Assign identification numbers
• Each element of the sample should have a unique
  identifier.
• Assure appropriate and desired level of
  measurement
• To ensure the use of selected statistical
  analysis, it is important that the item
  representing each variable be at the appropriate
  and desired level of measurement.

3
Preparation for Data Handling

• Establish and build coding procedures
• Assign numbers to categories
• Include spreadsheet column numbers on the
  instrument
• Check that the analysis is appropriate
• Check that each type of statistical analysis and
  summary of data yields data in the form you
  desire.

4
Preparation for Data Handling

• Plan for editing the data
• You can expect to find any or all of the
  following
• Nonresponse (Missing data)
• Incorrect responses
• Illegible handwriting
• Invective or strong language
• Inconsistent responses
• Plan time to examine data

5
Plan for data reduction and summaries

• Frequency distributions
• An arrangement of scores or categories with the
  frequency of occurrence (f) of each score or
  category indicated.
• Measures of Central Tendency
• The mean (requires interval)
• The median (requires median)
• The Mode (requires mode)

6
Plan for data reduction and summaries (contd)

• Measures of variability
• Variability degree to which scores within a
  distribution of scores differ from each other
• Variance s2, ?2
• Standard deviation s, ?
• The Range High-Low
• Contingency Tables and Cross tabulations
• A table which shows joint classifications of two
  or more variables, indicating the number (or )
  of observations in each of the joint cells.