|
|
More with our
Video examples
|
|
|
Return
|
|
|
|
|
Coding variable
With Excel you can easily create new variable but SurveyMiner adds coding procedures that Excel can’t do simply: coding numeric variables in classes with equal ranges or with equal frequency classes or user defined classes, coding categorical (multiple responses or not) variables in binary variables, grouping items, subtotals, new dimensions, weighting, automatic script creation. Scripts are used to save your coding formula and to create automatic coding procedures.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Weighting
When the structure of a particular sample analyzed is not
representative of the total population, weighting allows you
to assign a weight to each individual in order to correct the effect of
certain groups’ under-representation or over-representation in the sample.
Weighting can be done up to 6 criterion variables.
|
|
|
|
|
|
|
|
Sampling
With random sampling there is a random selection of individuals
from the total population. With quota sampling a constraint is
added to the random selection process. The resulting sample must
match a particular structure on one to six criterion variables.
|
|
|
|
|
|
|
|
Data Cleaning
Before you begin to analyze your data, be sure that the valid
responses are correct with the Cleaning data module. A data set
can contains some errors. To avoid drawing wrong conclusions, the
errors have to be first detected. Unless the dataset is small (i.e.,
less than 100 cases and 10 variables), data cleaning permits you to
verify that your data values are correct. Responses must fall into the
range of possible answers. You can also create a set of conditional rules
to find the errors: (i.e., young interviewees cannot be retired).
SurveyMiner change the improper values into missing values. Data cleaning
scripts are useful to automatically launch the same procedure
on different data sets.
|
|
|
|
|
|
|
|
Return
|
|