Small error in Philippines Raw Service Availability dataset [message #13148] |
Thu, 28 September 2017 14:51  |
Craig
Messages: 3 Registered: September 2017 Location: New York, NY
|
Member |
|
|
I'm specifically looking at PHSQ30FL.SAS, within the PHSQ30FL.ZIP file.
This is an SAS program, designed to interpret the .DAT flat file. Within this program, it assigns the characters in bytes 61-65 of each record to field S107 ("Transport to go to urban centr"), and then tells the program to interpret those bytes using the data dictionary for format S107_.
The data dictionary for S107_ reads:
value S107_
1 = "A.Walking"
2 = "B.Personal vehicle"
3 = "C.Hired vehicle"
4 = "D.Public transport"
5 = "E.Other"
6 = "?.Missing"
;
That is, it can interpret values "1","2",...,"6" as individual characters appearing in the record and treats the question as multiple-answer. However, when you look at the actual records, you find values "A",...,"E","?" appearing.
That is, the data dictionary would be correct if it read
value S107_
"A" = "Walking"
"B" = "Personal vehicle"
"C" = "Hired vehicle"
"D" = "Public transport"
"E" = "Other"
"?" = "Missing"
;
but as it stands it does not parse the records correctly.
Just letting you know; it's a relatively easy fix.
|
|
|