Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

From |
Eric Booth <ebooth@ppri.tamu.edu> |

To |
"<statalist@hsphsun2.harvard.edu>" <statalist@hsphsun2.harvard.edu> |

Subject |
Re: st: Generate a variable using Freq. of one variable and Percent of another variable |

Date |
Wed, 20 Jul 2011 20:27:23 +0000 |

<> Try this example in a do-file: *********************! clear **gen fake data q4 and q5: set obs 450 forval n = 4/5 { g q`n'a = int(runiform()*8)+2 g q`n'b = int(runiform()*7)+3 ta q`n'* } **create string proportion variable: forval n = 4/5 { g q`n'_num = . **prop of b answered each a answer choice- levelsof q`n'b, loc(val) foreach v in `val' { egen q`n'_`v'a = count(q`n'a) if q`n'a==`v' egen q`n'_`v'b = max(q`n'_`v'a) replace q`n'_num = q`n'_`v'b if `v'==q`n'a drop q`n'_`v'a q`n'_`v'b } recode q`n'_num (.=0) count if !mi(q`n'b) g q`n'_num2 = string(q`n'_num) + " / `r(N)'" } l in 1/10 ********************! - Eric __ Eric A. Booth Public Policy Research Institute Texas A&M University ebooth@ppri.tamu.edu \ On Jul 20, 2011, at 3:03 PM, Jonathan Grondin wrote: > Stata/MP 10.1 for Windows > Born 10 Jun 2010 > ado - 10 Jun 2010 > > > Hi there, > > I have conducted a questionnaire and every question can be answered by a value of 1 through 10. I have two groups who have taken the questionnaire, group "a" and group "b" and I have divided every question into 2 variables. > Question four would be divided this way: "q4a" and "q4b". > For every student and every question, I wish to compare the proportion of student from one group who provided the same answer as the students in the other group. > For example: Sam, a student in group "a", has answered "6" to question four. In group "b", out of ten students, three student have answered "6" to the same question. Thus, Sam's proportion for question four would be "3/10". > > Now, I want to generate a variable that would hold these proportions as values. > > > Here is a tabulate of the two groups on question four to explain better my need. Following, I have provided an example of the variable I would like to generate. > > > . tab q4a > > q4a | Freq. Percent Cum. > ------------+----------------------------------- > 1 | 12 2.78 2.78 > 2 | 72 16.71 19.49 > 3 | 109 25.29 44.78 > 4 | 122 28.31 73.09 > 5 | 67 15.55 88.63 > 6 | 26 6.03 94.66 > 7 | 9 2.09 96.75 > 8 | 8 1.86 98.61 > 9 | 4 0.93 99.54 > 10 | 2 0.46 100.00 > ------------+----------------------------------- > Total | 431 100.00 > > > . tab q4b > > q4b | Freq. Percent Cum. > ------------+----------------------------------- > 2 | 2 0.78 0.78 > 4 | 7 2.72 3.50 > 5 | 18 7.00 10.51 > 6 | 25 9.73 20.23 > 7 | 31 12.06 32.30 > 8 | 37 14.40 46.69 > 9 | 48 18.68 65.37 > 10 | 89 34.63 100.00 > ------------+----------------------------------- > Total | 257 100.00 > > > > I would like to get a variable like this > > > > . tab q4a > > q4a | Freq. Percent Cum. > ------------+----------------------------------- > 0 | 12 2.78 2.78 > 2/257 | 72 16.71 19.49 > 0 | 109 25.29 44.78 > 7/257 | 122 28.31 73.09 > 18/257 | 67 15.55 88.63 > 25/257 | 26 6.03 94.66 > 31/257 | 9 2.09 96.75 > 37/257 | 8 1.86 98.61 > 48/257 | 4 0.93 99.54 > 89/257 | 2 0.46 100.00 > ------------+----------------------------------- > Total | 431 100.00 > > > Any suggestions ? > > Thank you so much > > > Jonathan > * > * For searches and help try: > * http://www.stata.com/help.cgi?search > * http://www.stata.com/support/statalist/faq > * http://www.ats.ucla.edu/stat/stata/ * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/

**References**:**st: Generate a variable using Freq. of one variable and Percent of another variable***From:*Jonathan Grondin <jomatrix65@hotmail.com>

- Prev by Date:
**AW: st: on data transformation** - Next by Date:
**AW: st: heteroskedasticity and autocorrelation in GLS** - Previous by thread:
**st: Generate a variable using Freq. of one variable and Percent of another variable** - Index(es):