Information

Scottish Employer Perspectives Survey 2021: Technical Report

Technical report for the 2021 Scottish Employer Perspectives Survey.


Appendix G: Edit specification

Table 18: Specification of edits applied to the data
Edit number Objective of edit Condition Force/actions
1 Check how sectors given at A8 correspond to sample sector If A7 = 2 and sector6 different to samvars_sector6 Explore patterns of high frequencies of discrepancies between coded and sectors on sample.
2 To check that public sector ('local or central government financed body') definition given at A5 is consistent with SIC code "If A5 is '3' or '4' and SIC2 is not '84', '85' or '86' To check that public sector ('local or central government financed body') definition given at A5 is consistent with SIC code
3 To check legitimacy of high numbers of employment If A1 is greater than 1500 Each record to be judged on an individual basis, using internet searches where necessary. Call-backs may be necessary if number of employees seems excessively high for business activity.
4 To check how A1 answers match to Sample size If A1 is more than 20 times samvars_emps if A1<10 OR If A1 is more than 15 times greater or less than samvars_emps if A1 IS BETWEEN 10 AND 99 OR If A1 is more than 10 times greater or less than samvars_emps if A1 ≥ 100 To check how A1 answers match to Sample size
5 To check plausibility of history / numbers of employment If A1 ≥ 50 and F1=1 Each record to be judged on an individual basis, using internet searches where necessary. Consider appropriate action if discrepancies are inexplicable.
6 Check that all SICs given at A8 are valid SIC07 codes If A7 = 2 Any records where A7=2 will need to have the SIC at A8 validated. Recode where possible, recontact if necessary.
7 To ensure that all SOC codes are valid at C8 If SOC at C8 is not in full SOC file provided Recode where possible, recontact if necessary.
8 To check high numbers on particular work experience placements If C18_1-11 is (combined) more than 5 times greater than A1 if A1<5 OR If C18_1-11 is (combined) more than 2 times greater than A1 if A1 IS BETWEEN 5 AND 49 OR If C18_1-11 is (combined) greater than A1 if A1 ≥ 50 OR If any C18_1-11 value greater than 100 Each record to be judged on an individual basis.
9 To check high numbers of apprentices If D22a is more than 5 times greater than A1 if A1<5 OR If D22a is more than 2 times greater than A1 if A1 IS BETWEEN 5 AND 49 OR If D22a is greater than A1 if A1 ≥ 50 OR If D22a is greater than 100 Each record to be judged on an individual basis.
10 To check that extensive use of initiatives is plausible If count of number of initiatives/schemes used across C3, D2 and D5 is >10 Assess responses to other questions for engagement with the external training landscape. Consider appropriate action if amount feels unreasonable.
11 To check contradictory responses on how long establishment has been operating and when offered Apprenticeships If F1=1 and D23i=1 (offered Apps in last 3 years) Review on an individual basis. Consider data edits at D23i or A6 if contradictions are present.
12 Check legitimate use of the "Head Office" code If establishment is a single site (A2=1) and has used one of the various 'Head Office Decision' codes at C27, D27c, D27ai, D33, D34a. Remove use of the Head Office code if establishment is a single site.
13 Check for excessive use of 'Don't know' code at key measures If answered "Don't know" at C1, C6A, D6A, D6B, D13 and D22. Each case to be reviewed and act as a sense check that numbers that fail this check are low. (To be 'DK' across so many key measures would be indicative of possibly having not interviewed the most appropriate respondent).

Contact

Email: FHEstatistics@gov.scot

Back to top