Data Cleaning

[KANISHKA SUBASHINI MANI]

Hi,

My variable region looks like this and I'm not able to find the reason for each category repeating itself. How would you suggest I should take this forward.

[Daniel Schaefer]

It looks like the label repeats itself, but the value of the variable does not. If I were you, I might start by returning to the codebook or other data documentation and trying to figure out what the labels should be for each value. You can see the value (instead of the label) with the following line:

tab region, nolabel

[Chen Samulsion]

The -fre- command witten by Ben Jann will do this job #2 better.

ssc install fre
fre region

[KANISHKA SUBASHINI MANI]

Thank you Daniel Schaefer & Chen Samulsion

[KANISHKA SUBASHINI MANI]

How should I proceed after this?

[Nick Cox]

I think Daniel Schaefer was on target here. The answer lies in how these data were produced. But there is an easier route to follow. A leading space in some cases but not others would account for your results.

. clear

. set obs 4
Number of observations (_N) was 0, now 4.

. gen foo = word("Central Eastern Central Eastern", _n)

. tab foo

        foo |      Freq.     Percent        Cum.
------------+-----------------------------------
    Central |          2       50.00       50.00
    Eastern |          2       50.00      100.00
------------+-----------------------------------
      Total |          4      100.00

. replace foo = " " + foo in 3/4
variable foo was str7 now str8
(2 real changes made)

. tab foo

        foo |      Freq.     Percent        Cum.
------------+-----------------------------------
    Central |          1       25.00       25.00
    Eastern |          1       25.00       50.00
    Central |          1       25.00       75.00
    Eastern |          1       25.00      100.00
------------+-----------------------------------
      Total |          4      100.00

. fre foo

foo
-------------------------------------------------------------
                |      Freq.    Percent      Valid       Cum.
----------------+--------------------------------------------
Valid   Central |          1      25.00      25.00      25.00
        Eastern |          1      25.00      25.00      50.00
        Central |          1      25.00      25.00      75.00
        Eastern |          1      25.00      25.00     100.00
        Total   |          4     100.00     100.00          
-------------------------------------------------------------

In my example fre is showing this space -- and so was tabulate. It is just hard to see.

For leading space, read also any space-like characters. If

replace region = trim(region)

does not solve the problem, then find out which characters are present using chartab from SSC.