Descriptive Statistics where the Data Type is Text?

I have a data sheet where all of the variables are descriptors, e.g., left lateral recumbency, ventral, prone, etc. I went to Stat>Basic Statistics>Display Descriptive Statistics, and none of the column names were in the dialog box. How do I get counts for each of the variables? Do I have to change them to numeric values?

Thanks.

Re: Newbie: Descriptive statistics where the Data Type is Text?

You do not specify the specific statistics that you want to obtain. If you are strictly interested in counts and percentages, you might try the Stat > Quality > Pareto chart routine.

Re: Newbie: Descriptive statistics where the Data Type is Text?

Miner said:

You do not specify the specific statistics that you want to obtain. If you are strictly interested in counts and percentages, you might try the Stat > Quality > Pareto chart routine.

Click to expand...

For starters, I merely want to know how many subjects were in left lateral recumbency, right lateral recumbency, standing,etc. There is one column for "posture", but because it is in text and not numbers, I can't get the counts in the descriptive statistics.

Thank you for responding!

Re: Newbie: Descriptive statistics where the Data Type is Text?

Descriptive statistics in Minitab are only available for numeric variables. Therefore they are not shown in the variables field obtained with Stat > Basic Statistics > Show (or Store) Descriptive Statistics. Metrics like mean, standard deviation and so on couldn't give you helpful informations about text variables, so Minitab don't show these.

To get percentages and frequencies, use

Stat > Tables > Tally Individual Variables
(for 1 text variables)

Stat > Tables > Cross Tabulation and Chi-Square
(for 2 or more text variables)

Regards,

Barbara

Re: Newbie: Descriptive statistics where the Data Type is Text?

I always find Stat>Tables>Tally, as Barbara mentions really useful for these investigations.
There are also Stat>Tables>Descriptive statistics
This allows some interesting layering of categorical factors as well.
If you want to use this data graphically Bar Charts will, from Counts of Unique values, present this quite nicely.