Thanks for checking.

I was expecting the missing values to appear either at the top or bottom of the list for numeric variables. In my dataset, it appears first for variables with 2-3 values, but in variables with greater than 3 values, it appears in an arbitrary position.

I don't see how the *.chk file could affect (which controls entry) could affect the position of . in the codebook since that tabulation occurs later after the file is created.

RH/NIH

epidata-list@lists.umanitoba.ca wrote: I gave it a try: - When the variable is NUMERIC (as it is the case in your file), missing (.) values appear at the end of the tabulation. When it is STRING, missing (.) values appear at the top of the tabulation as follows: var1S -------------------------------------------------------------------- var1S type: Texte majuscule Libellés des codes: libellé_var1s

manquant: 3/7 valeurs uniques: 4

tabulation: Freq. P.100 Valeur Libellé 3 42.9 . 1 14.3 1 1 1 14.3 2 2 2 28.6 3 3

var2N -------------------------------------------------------------------- var2N type: Numérique Libellés des codes: libellé_var1s

manquant: 3/7 fourchette: [2 ; 3] valeurs uniques: 3

tabulation: Freq. P.100 Valeur Libellé 3 42.9 2 2 1 14.3 3 3 3 42.9 .

Note: as shown in your example, there is a redundancy in the checkblock: the variable is controlled by a list if valid values ('label_ramde7'), AND there a Range command (0-10,98). As it is already controlled by the list, it is not useful to control it twice with 'Range - legal'. Choose either of the 2 control systems, remove the other one, and check again.

______________________________________________ Gilles DELMAS Institut de Veille Sanitaire Dept. Maladies Infectieuses, Unité infections entériques, alimentaires et zoonoses 12 rue du Val d'Osne 94415 Saint-Maurice cedex - France + 33 1 41 79 67 27 g.delmas@invs.sante.fr ______________________________________________

-----Message d'origine----- De : epidata-list@lists.umanitoba.ca [mailto:epidata-list@lists.umanitoba.ca] Envoyé : mardi 28 mars 2006 00:00 À : epidata-list@lists.umanitoba.ca Objet : RE: [EpiData-list] second try: order of missing valuesindocument->codebook

That variable is a 2 digit number created with ## in the qes file.

Here's the rest of the info about the variable in the codebook:

type: Number value labels: label_ramde7 range/legal: 0-10,98

missing: 6/52 range: [3 ; 10] unique values: 8

epidata-list@lists.umanitoba.ca wrote: What kind of variable is it? This shouldn't happen id the variable is numeric.

-----Message d'origine----- De : epidata-list@lists.umanitoba.ca [mailto:epidata-list@lists.umanitoba.ca] Envoyé : vendredi 24 mars 2006 23:40 À : epidata-list@lists.umanitoba.ca Objet : [EpiData-list] second try: order of missing values indocument->codebook

Didn't hear from anyone, so I'm trying again (sorry).

I have been creating a dataset using epidata. I was looking through the documentation->codebook to get a sense of the distribution ofvalues for thevariables. I just noticed that the missing values (.) appear at an almost random point in the otherwise ordered list (in this example,between 9 and 10). I compared the codebook with SAS tables and the data appear to be correct.

tabulation: Freq. Pct. Value Label 2 3.8 3 1 1.9 4 2 3.8 5 6 11.5 7 7 13.5 8 7 13.5 9 6 11.5 . 21 40.4 10

Any ideas why this might happen?

thanks.

Richard Herrell NIH

--------------------------------- Yahoo! Messenger with Voice. PC-to-Phone calls for ridiculously low rates. _______________________________________________ EpiData-list mailing list EpiData-list@lists.umanitoba.ca http://lists.umanitoba.ca/mailman/listinfo/epidata-list

--------------------------------- Blab-away for as little as 1¢/min. Make PC-to-Phone Calls using Yahoo! Messenger with Voice. _______________________________________________ EpiData-list mailing list EpiData-list@lists.umanitoba.ca http://lists.umanitoba.ca/mailman/listinfo/epidata-list

--------------------------------- Yahoo! Messenger with Voice. PC-to-Phone calls for ridiculously low rates.