| Scottish Household Survey | Scottish House Condition Survey (SHCS) |
|---|
National Statistic | YES | No |
Contractor | MORI and TNS | Office for National Statistics and Miller Mitchell Burley Lane |
Contract Length | Jan 2003 - Dec 2006 | 2003 - 2006 (new contracted to be awarded in Spring 2006) |
Contract Cost | £4M | £5m approx |
Start of next tender process | Spring 2006 | November 2005 |
Period of next potential contract | July/August 2006-July/August 2011 Fieldwork period: January 2007-December 2010 | 2006 - 2009 |
Survey Reference Period | Calendar years | October to September |
Survey Structure | Continuous and cross-sectional | Continuous and cross-sectional |
Interview Mode and specialised components | Face-to-face CAPI | CAPI interview Building survey |
Sampling Frame | Postal Address File | Postal Address File |
Sampling Design | Systematic random sample where fieldwork conditions allow (areas of high population density). Unclustered in local authorities where population density of 500 or more persons per square kilometre. Clustered in those local authorities where population density of 500 persons or less per square kilometre. | Stratified random sample designed to give minimum number of paired surveys per local authority area |
Primary reporting units | Households and random adults | dwellings |
Secondary reporting units | Children | households |
Response rate | 1999/2000 - 66% 2001/2002 67% 2003/2004 69% | 70% response to social survey 83% conversion rate to physical survey |
Target Sample size (households) | 15.5K per calendar year (31,000 over the two year sweep) | 3000 paired (social and physical) cases per year |
Target Sample size (random adults) | Target is to complete random adult interview in 90% of households in which a household interview is carried out. | n/a |
Target Sample size (children) | We don't have an explicit target sample size for any attribute other than the expectation that a random sample with a high response rate will yield a sample with attributes that reflect the population in private households with the limits of sampling variability. | n/a |
Door step routine | Interviewers are required to make up to six calls at an address (an initial visit plus five 'call-backs'). In addition to the immediate reissue of contact sheets that have been wrongly completed or where the required number of call-backs has not been made, there is an on-going programme of reissuing 'non-contacts' in a bid to maximise the response rate. At the end of each fieldwork year a significant number of valid but 'non-contact' addresses remain 'live'. | Interviewers make minimum of six calls (at least two at evening and weekends) and use kish grid to select households where there are multiple addresses. Physical surveyors make four visits - at least two at evenings and weekend |
Field force size | Each month there are about 100 interviewers working on the SHS from the two companies. Not all will work each month so over the course of a year, approximately 200 (214 in 2004) will work on the SHS. | ONS employ approx 60 interviewers on SHCS - MMBL approx 50 surveyors |
Groups excluded from sample | Communal establishments | Communal establishments |
Sample coverage | Scotland | Scotland |
Questionnaire quality assurance | Individual questions are discussed and agreed between the contractor and the Executive. These are rarely subject to piloting (the only example I can think of is the volunteering questions which were piloted although not by us). They are often drawn from other government surveys so have been piloted in that respect. The research team on the contractor side specify the questions for scripting. These are checked by the person responsible for the questions within the SE. Scripted questions are checked on paper - reading them and the associated programming - and comparing against spec. The full CAPI script is tested by a number of researchers in the contractor team, with each route in the questionnaire checked to ensure that the script works as specified. Checking is generally limited by the fact that questions tend to be agreed very close to the point where the questionnaire needs to go into the field. Early in the fieldwork, data is extracted from the CAPI servers and checked to ensure that the scripted questions are working properly. | Hard and soft checks built in to CAPI script - validation of physical survey is done by MMBL and Communities Scotland |
Dataset quality assurance | Most of the quality assurance on the data is built into the script, which checks the logic of responses, especially in the HA section, which collects age, sex, relationship and occupational data for each household member. These checks are confirmed as part of the data processing. In addition, checks are carried out on all data as part of the process of publication. | Data validation carried out by ONS and Communities Scotland |
Frequency of data delivery from contractor | Quarterly | Quarterly |