Common Questions:
Number of observations (rows):
Generally speaking, the more the better
At least 15 observations
Feel free to cut your original sample if it is too big
Examples: each row is for one year, one individual or a combination of year and individual
Dataset Type 1: time series (the same individual for multiple time periods)
Dataset Type 2 : cross-sectional data (multiple individuals for the same time period)
Dataset Type 3: panel data (multiple individuals for multiple time periods)
Number of variables of interest (columns):
At least 2
Variable type:
Numerical: use as they are or create categories
Categorical: create binary variables if more than two categories
For example, eye color: blue(1), green(2), brown(3), others (4). To include eye color in your regression as an independent variable, you need to create four binary variables,
e.g. Blue = 1 if eye color is blue; 0 otherwise;
Green = 1 if eye color is green; 0 otherwise;
Brown = 1 if eye color is brown; 0 otherwise;
Others = 1 if eye color is not blue, green or brown; 0 otherwise.
Then you include three out of the four binary variables as your independent variables.