Weighting Methodology for the Private Landlords Survey Robert Bucknall, ONS.

Slides:



Advertisements
Similar presentations
Multiple Indicator Cluster Surveys Survey Design Workshop
Advertisements

Multiple Regression.
Disability and pay: a decomposition of the pay gaps of disabled men in the UK Simonetta Longhi, Cheti Nicoletti and Lucinda Platt ISER, University of Essex.
Sampling.
Depression and work incapacity in Scotland: Evidence from the Scottish Health and British Household Panel Surveys Matt Sutton Will Whittaker Health Methodology.
Outline of talk The ONS surveys Why should we weight?
Longitudinal LFS Catherine Barham and Paul Smith ONS.
Family Resources Survey Proposals for the treatment of unlinked FRS data* Valerie Christian & Philip Clarke.
Online Privacy Survey Results Conducted: December 2011.
Statistical Issues in Measuring Poverty from Non-Survey Sources NATIONAL ACCOUNTS UNSD/NA/MR1 UN STATISTICS DIVISION Economic Statistics Branch National.
Household Projections for England Yolanda Ruiz DCLG 16 th July 2012.
©2013 Experian Limited. All rights reserved. Experian and the marks used herein are service marks or registered trademarks of Experian Limited. Other products.
Sociology 680 Multivariate Analysis Logistic Regression.
An Assessment of the Impact of Two Distinct Survey Design Modifications on Health Insurance Coverage Estimates in a National Health Care Survey Steven.
‘OPPORTUNITY KNOCKS ’ FOR THE PRIVATE RENTED SECTOR JOHN MASON HEAD OF POLICY & COMMUNICATION.
Microsimulation of Survey Collection Yves Bélanger Kristen Couture 26 January 2010.
Adjustments for Age-sex and MLC NRAC 29 March 2007.
Life Opportunities Survey (LOS) Wave 2 Weighting Andy Fallows and Sangeetha Gallagher.
Weighting and Imputation for CORE Social Housing Statistics Julia Bowman & Niall Goulding.
Small Area Estimates of Fuel Poverty in Scotland Phil Clarke (ONS), Ganka Mueller (Scottish Government)
Sample of Anonymised Records: User Meeting Propensity to migrate by ethnic group: 1991 & 2001 Paul Norman 1, John Stillwell 2 & Serena Hussain 2 School.
QUALITATIVE AND LIMITED DEPENDENT VARIABLE MODELS.
BACKGROUND RESEARCH QUESTIONS  Does the time parents spend with children differ according to parents’ occupation?  Do occupational differences remain.
Change in prevalence of Chronic Kidney Disease in England over time: comparison of nationally representative cross-sectional surveys from 2003 to 2010.
1 Health Status and The Retirement Decision Among the Early-Retirement-Age Population Shailesh Bhandari Economist Labor Force Statistics Branch Housing.
Attitudes to the private rented sector in Ireland: Landlord and tenant survey results ENHR, London March 2015 David Duffy, ESRI.
Vulnerabilities in a Recovering Market: Experiences of Low Income Tenants in the PRS ENHR Private Rented Markets Seminar 20 th March 2015.
Joint UNECE/Eurostat Meeting on Population and Housing Censuses (13-15 May 2008) Sample results expected accuracy in the Italian Population and Housing.
Housing Studies Association 2015 Dr Alison Wallace BUY-TO-LET MORTGAGE ARREARS: Understanding the factors that influence landlords’
Household projections for Scotland Hugh Mackenzie April 2014.
Voting Behavior of Naturalized Citizens: Sarah R. Crissey Thom File U.S. Census Bureau Housing and Household Economic Statistics Division Presented.
National Household Survey: collection, quality and dissemination Laurent Roy Statistics Canada March 20, 2013 National Household Survey 1.
Factors that Associated with Stress in Nursing Faculty in Thailand
The new HBS Chisinau, 26 October Outline 1.How the HBS changed 2.Assessment of data quality 3.Data comparability 4.Conclusions.
18 September Health Plan Actuarial Value Variation Among Employers Actuarial Research Corporation Sarah Yi Jim Mays Middle Atlantic Actuarial Club.
Definitions Observation unit Target population Sample Sampled population Sampling unit Sampling frame.
The effect of uncertainty on fuel poverty statistics Laura Williams, Department of Energy and Climate Change GSS Methodology Symposium, 6 th July 2011.
Evidence-Based Medicine 3 More Knowledge and Skills for Critical Reading Karen E. Schetzina, MD, MPH.
Land Rental Markets in the Process of Structural Transformation: Productivity and Equity Impacts in China Songqing Jin and Klaus Deininger World Bank.
Use of Administrative Data in Statistics Canada’s Annual Survey of Manufactures Steve Matthews and Wesley Yung May 16, 2004 The United Nations Statistical.
Hampshire, Portsmouth & Southampton Home Movers Survey 2010 PRLG 22 nd September 2010.
Emerging methodologies for the census in the UNECE region Paolo Valente United Nations Economic Commission for Europe Statistical Division International.
Welfare Reform and Lone Parents Employment in the UK Paul Gregg and Susan Harkness.
Scot Exec Course Nov/Dec 04 Survey design overview Gillian Raab Professor of Applied Statistics Napier University.
3.14 X AXIS 6.65 BASE MARGIN 5.95 TOP MARGIN 4.52 CHART TOP LEFT MARGIN RIGHT MARGIN © TNS 2014 Fieldwork effort, response rate and the distribution.
Americas Desk OECD Development Centre LAC Fiscal Policy Forum Panama, September 16 th 2010 Fiscal policy in Latin America: Fiscal legitimacy and net tax/benefit.
Improving the Quality of the HMRC Personal Wealth Statistics Rebecca Ambler and Abeda Malek - HMRC.
Using administrative registers in sample surveys European Conference on Quality in Official Statistics 3-–6 May 2010 Kaja Sõstra Statistics Estonia.
Chapter 15 Sampling and Sample Size Winston Jackson and Norine Verberg Methods: Doing Social Research, 4e.
Things that May Affect the Estimates from the American Community Survey Updated February 2013.
5-4-1 Unit 4: Sampling approaches After completing this unit you should be able to: Outline the purpose of sampling Understand key theoretical.
Updating Household Projections for England Bob Garland.
Chapter 6: 1 Sampling. Introduction Sampling - the process of selecting observations Often not possible to collect information from all persons or other.
AN EXAMPLE OF COOPERATION & SOME WIDER ISSUES Ian Plewis (Bedford Group, Institute of Education) & Stephen Morris (Social Research Division, Department.
Improving of Household Sample Surveys Data Quality on Base of Statistical Matching Approaches Ganna Tereshchenko Institute for Demography and Social Research,
Modelling international migration to produce local level estimates Ruth Fulton Office for National Statistics.
Interviewer Effects on Paradata Predictors of Nonresponse Rachael Walsh, US Census Bureau James Dahlhamer, NCHS European Survey Research Association, 2015.
Changes to the collection of short walk data in the NTS Glenn Goodman, DfT.
1. Divide the study sample data into two groups: Fatigued (F), N=3,528, and Non-Fatigued (NF), N=3,634. Estimate logistic regressions to obtain the probability.
An ecological analysis of crime and antisocial behaviour in English Output Areas, 2011/12 Regression modelling of spatially hierarchical count data.
Analysis of the characteristics of internet respondents to the 2011 Census to inform 2021 Census questionnaire design Orlaith Fraser & Cal Ghee.
JP Research Summary: JASIC Flex Injury Estimate GTR09 PH2 Informal Working Group 17 Sept GTR
Statistics Canada Citizenship and Immigration Canada Methodological issues.
Q2010 – special topic session 33 - Page 1 Indicators for representative response Barry Schouten (Statistics Netherlands) Natalie Shlomo and Chris Skinner.
Guillaume Osier Institut National de la Statistique et des Etudes Economiques (STATEC) Social Statistics Division Construction.
IAOS Shanghai – Reshaping Official Statistics Some Initiatives on Combining Data to Support Small Area Statistics and Analytical Requirements at.
Presented by: Khaleel S. Hussaini PhD Bureau Chief, Public Health Statistics Division of Public Health Preparedness Judy Bass Arizona’s BRFSS Coordinator.
Inflation Report February Output and supply.
Evaluating imputation of sex and age for substitutes in substitute households Michael Ryan 2008 UNECE Work Session on Statistical Data Editing.
Salah Merad Methodology Division, ONS
Presentation transcript:

Weighting Methodology for the Private Landlords Survey Robert Bucknall, ONS

Presentation outline Overview of Private Landlords Survey Potential sources of bias Weighting of the Survey Dwelling weights Landlord weights Calibration Recommendations

Overview of the PLS Follow up to English Housing Survey Collects information about: Ownership Occupation Management Practices of privately rented dwellings in England No available list of private landlords from which to conduct survey. Based on sample of landlords derived from the EHCS and EHS

EHCS/EHS Questions about Landlord

PLS sample selection

Dwelling and Landlord weights Dwelling derived sample Previously, distributions reported always related to % of dwellings rather than % of landlords Became issue as policy focus of the survey shifted from dwelling condition and maintenance to landlord letting and management practices

Dwelling based output table Example of dwelling based table from 2006 PLS:

Landlord based output table Example of landlord based table from 2006 PLS:

Stage 1 weight A response model was developed: For each dwelling on the EHCS/EHS private renters dataset, a PLS response marker was created: »1 if on PLS sample »0 if not on PLS sample A survey weighted logistic regression was conducted to model the likelihood of receiving landlord details from each dwelling

Stage 1 weight

The following variables were considered as predictors in logistic regression: Region Ethnicity of household reference person Housing benefits Property type Satisfaction with the service provided by the landlord Furnished or unfurnished Employment status of household reference person Age of household reference person Marital status

Stage 1 weight The logistic regression model identified the following variables as predictors of a tenant’s propensity to provide their landlord’s contact details: Region Housing benefits Satisfaction with the service provided by the landlord Age of household reference person Marital status

Stage 1 weight The probability of receiving landlord contact details from tenant i is given by the logistic function: where are regression coefficients and are explanatory variables for tenant i.

Stage 1 weight The Stage 1 weight for dwelling i on the PLS is: where is the EHCS/EHS weight for dwelling i and is the probability that landlord contact details are provided by dwelling i

Stage 2 weight (landlord weight) Landlords who own large portfolios of properties have a greater chance of being included on the frame than landlords who own small portfolios of properties Analysis has shown that there is significant variation in landlord response rates across region and landlord portfolio size

Stage 2 weight (landlord weight)

The landlord weight should reflect the probability of the landlord being sampled A dwelling adjustment reflects the size of the portfolio of tenant i’s landlord: where is the Stage 1 weight for dwelling i and is the portfolio size of landlord L

Stage 2 weight (landlord weight) The initial weight for landlord L is: where is the dwelling adjustment and is the Stage 1 weight for dwelling i

Stage 2 weight (landlord weight)

Adjustments required to minimise landlords non- response bias 18 weighting classes (9 GOR x 2 sizebands) Landlord weights were adjusted to account for landlord non-response: where is the number of sampled landlords in GOR g sizeband z, and is the number of responding landlords in GOR g sizeband z

Calibration Estimates of private renters within GOR available from LFS PLS landlord weights calibrated to LFS private renter estimates GES ensures weights sum to predetermined totals

Calibration Initial landlord weights and portfolio sizes Control variables (LFS household totals by GOR) Final landlord weights

Recommendations New PLS weighting method minimises potential bias in the survey: Tenant’s reluctance to provide landlord contact details Over representation of larger landlords on the PLS Loss of landlords due to non-response to PLS Method for producing landlord weights developed due to a shift in policy focus of the survey

Questions / Comments