NPLH GSS Data Cleaning Log 9/20/13 TX Deleted two duplicates

advertisement

NPLH GSS Data Cleaning Log

9/20/13 TX Deleted two duplicates: (in syntax)

1.

SM ID 1997750229 (employee ID 154146): deleted the one that was incomplete

2.

SM ID 2516079906 (employee ID 186036): deleted the one filled in on a later date (3/14/13).

9/20/13 TX Recoded Race_Other to Race_WH (in syntax)

1.

Race_Other as Hispanic/Mexican/Latino with answering Hispanic ethnicity to Race_WH (SM ID

1997834269, 1997784535, 1997792702, 2000259644, 1997192156, 2000286014, 1995949869,

1997832224)

2.

Race_Other as Romanian to Race_WH (SM ID 1995985436)

10/4 TX VAR Other language: Employee ID 2084000, “English” as skipped question

Does this mean Bilingual was set to No and Non_english_language was set to “”?

10/4 TX County _update CNTY from “both” to: employee ID 130055 to Dallas, employee ID 169328 to

Tarrant, based on a review of meeting logs that indicated they had not worked in both counties. (in syntax)

11/8 CO Recode Race_Other to Race_WH (in syntax)

1.

Race_Other as Latino/Hispanic to Race_WH (SM ID 1994887445, 2582842996) (also needed to make Hispanic 1 for first ID)

2.

Delete Race_Other and Keep Race_WH (SM ID 1995998254, 1996005505) (Answered American,

Irish…)

3.

Delete Race_Other (answered brown) (SM ID 1995997149)

11/11/13 SD Recoded Current_Job_X

1.

Respondents who selected Other and then indicated their current job involved family group coordination were recoded from “Other” to “FGDM” (SM ID 2053181991, 2053182332,

2053182721) (in syntax)

11/19 CO duplicate surveys (in syntax, 12 in all)

1.

Employee with Emp_ID # 100002231 aka 10000 2231, completed the survey 11 times. We selected the earliest complete survey for use. The following SM IDs were deleted: 2783007897,

2797333555, 2699213335, 2724944052, 2731512632, 2741497674, 2741531397, 2754901642,

2765114798, 2784653743. That said, the respondent had a lot of feedback in her subsequent submissions about barriers and thus it will be helpful to look over the original Colorado data file when summarizing responses to this question.

2.

Emp_ID #141432 started the survey twice on the same day and only completed it once. The incomplete survey, associated with SM ID 2583012746, has been removed.

3.

Emp_ID #100014400 was associated with two SM ID’s: 1994887531 and 1996003520. One was completed on 9/11/12 and the other on 9/12/12. While some demographic data align (primary job, sex and ethnicity) many other variables don’t appear to align (e.g., years in CW = 2 in one

NPLH GSS Data Cleaning Log survey, 7 in another education level = 7 in one survey, 9 in another and she skipped entering her age on the first survey). After conferring with John, we opted to remove the first entry. Since some demographics do not align and the first survey has a number of n/a responses, and no comments and the second has comments and no n/a responses, we opted to keep the second entry under that ID number. Therefore SM ID 1994887531 was deleted.

11/20/13 CO

1.

Four staff entered names as opposed to numeric employee IDs. We need to identify their actual employee ID numbers. Their names are: calvilvx, Osbornvt, Shana Ryken, wichmajr. (in syntax)

2.

Recoded Primary job for SM ID2326240331 from other (after hours) to caseworker and mapped her position to Investigation/assessment under current_ job _x (in syntax)

3.

Current Job_x Other response recoding: a.

Recoded to FGDM: SM ID 2582867581 1996008487 1994398797 1996008988

1996010412 1996007077 1996014300 1996005505 1996006398 1994391021

2583027268 2586880015 (in syntax) b.

Recoded to investigation/assessment for SM ID 2326237274 2326239555 1996006131

2583016965 (in syntax) c.

Recoded to ongoing SM ID 2326233455 (in syntax)

1/7/14 Jason further cleaning/recoding other_job (in syntax)

1.

SM ID 1995997749 (CO emp 100006559) entered “ALL of the above” but apparently did not select all of the above  made Curr_job_x_INV and _ONG = 1 and _OTHER = 0

2.

SM ID 1994394441 (CO emp 100004521) similarly entered “All areas”  made Curr_job_x_INV and _ONG = 1 and _OTHER = 0

3.

SM ID 2053182197 (SD emp 312) entered “Supervision of Intake, IFA, and Ongoing”  made

Curr_job_x_INV and _ONG = 1 and _OTHER = 0

4.

SM ID 2053182464 (SD emp 63) entered “Also supervise FGC coordinators and Kinship

Specialist”  made Curr_job_x_FGDM = 1 (_ONG already = 1) and _OTHER = 0

1/27/14 Jason cleaning new TX responses (all in syntax)

Duplicates: as above, deleted later response

1.

SM ID 1994402967 and SM ID 3008058881: 1 st taken 9/11/12 under employee ID 00000130211,

2 nd employee ID 130221. Yes, those are off by 1 digit, but ethnicity, language, education, and experience responses are identical. In 1 st , age = 38; in 2 nd , 40…and she helpfully supplied her birthdate instead of today’s date on her 2 nd try: 10/3/73.

2.

Other repeats: SM ID 3013136514 (kept 1997740620), 3010379663 (kept 1997750796),

3010389921 (kept 2000249852)

NPLH GSS Data Cleaning Log

Other work area:

1.

SM ID 3010416789 stated “fbss includes investigation task”—already had FBSS but did not check

Investigation  made Current_job_INV=1 and _Other = 0

Other ethnicity

2.

SM ID 3010414939 entered “white and black”…but checked neither. Made Race_BL = Race_WH

= 1 and Race_Other = 0

3/18/14 fixing employee ID (in syntax)

1.

Change garzame2 to 245262

Download