Keyword search

Filter results by

Search Help
Currently selected filters that can be removed

Keyword(s)

Survey or statistical program

738 facets displayed. 0 facets selected.

Content

1 facets displayed. 0 facets selected.
Sort Help
entries

Results

All (26,768)

All (26,768) (1,430 to 1,440 of 26,768 results)

  • Articles and reports: 12-001-X202600100006
    Description: We introduce a general framework for constructing master samples that preserve desirable design properties across panels. The core procedure is to order an initial probability sample. Since the final sequence must be robust to a uniform random rotation, we define and minimize an objective that aggregates panel-level performance across all possible circular panels. A final random rotation is applied to ensure design validity. The framework is flexible with respect to the choice of design criteria, such as spatial balance or marginal balance, and can be implemented efficiently using simulated annealing to obtain high-quality approximate solutions. By construction, the approach supports both positive and negative sample coordination for spatially balanced, marginally balanced, and doubly balanced samples. The method’s versatility is demonstrated through three applications: constructing a master sample with spatially balanced panels, marginally balanced panels, and doubly balanced panels.
    Release date: 2026-06-29

  • Articles and reports: 12-001-X202600100007
    Description: National statistical institutes operate sample coordination systems to spread the response burden in business surveys. Despite the applied sample coordination and monitoring the response burden, some businesses might still be heavily sampled within a short period. This may lead to a peaking response burden for individual businesses, which could affect response rates and response quality. This paper proposes a new sample coordination method based on Adapted Spatially Correlated Poisson (ASCP) sampling that focuses on businesses with a high response burden. The effects on the response burden will be evaluated in two simulation studies and compared with a stratified approach, a pragmatic method in which sampling fractions are manually adjusted and with the baseline method of ignoring the response burden. For the simulations, real-world scenarios and data from Statistics Netherlands are used. The first simulation study considers a practical situation in which a given sample is adjusted with the aim to avoid the occurrence of businesses with a peaking response burden. The second simulation study analyzes the longer-term effects of the different sample coordination methods and focuses both on the reduction and spread of the response burden. The advantages and disadvantages of the different methods will be explained and discussed in detail, and recommendations for applying these methods at national statistical institutes and other survey agencies will be given.
    Release date: 2026-06-29

  • Articles and reports: 12-001-X202600100008
    Description: This paper introduces an innovative and intuitive finite population sampling method that has been developed using a unique graphical framework. In this approach, first-order inclusion probabilities are represented as bars on a two-dimensional graph. By manipulating the positions of these bars, researchers can create a wide range of different sampling designs. This graphical visualization of sampling designs facilitates the exploration of alternative designs and may simplify certain aspects of the implementation compared to traditional mathematical algorithms. This novel approach holds significant promise for tackling complex challenges in sampling, such as achieving an optimal design. By applying a version of the greedy best-first search algorithm to this graphical approach, the potential for integrating intelligent algorithms into finite population sampling is demonstrated.
    Release date: 2026-06-29

  • Articles and reports: 12-001-X202600100009
    Description: Combining estimates from independent surveys via inverse-variance weights can lead to negative bias when unknown variances are estimated and the target variable is non-negative and positively skewed. In such cases, strong positive correlations typically arise between the estimators and their corresponding variance estimators, causing standard linear combinations with inverse-variance weights to exhibit negative bias. We introduce a strikingly simple method to reduce bias: replace the standard weight with the ratio of the estimator to the variance estimator. Under a linear model linking the two, we show that the new ratio-weighted estimator is approximately unbiased, whereas the conventional inverse-variance combination exhibits downward bias. Through simulations, we demonstrate that the new method brings both the bias and the mean squared error closer to the optimum for a wide range of different target variables. As our method uses only standardly reported summary statistics, it can be immediately adopted to reduce this widespread bias and improve the reliability of scientific findings in various fields.
    Release date: 2026-06-29

  • Articles and reports: 12-001-X202600100010
    Description: With the exception of two-phase sampling, the standard variance approximation of the generalized regression (GREG) estimator assumes that the population totals in the weighting scheme are observed without error. If the weighting model of the GREG estimator contains population totals that are observed with measurement error sources other than the sampling error of first-phase estimates, then this uncertainty will be ignored by the variance approximation of the GREG estimator. This paper proposes a variance approximation for the GREG estimator that accounts for additional uncertainty arising from measurement error in one or more of the population totals used in the weighting scheme. This approach has been developed for, and is being applied to, the Dutch Labour Force Survey (DLFS). The monthly publications of the DLFS are obtained with a time series model, which corrects for rotation group bias and discontinuities caused by major redesigns and the loss of face-to-face interviews during COVID-19. The GREG estimates for the quarterly figures are benchmarked to the average of the monthly publications to enforce numerical consistency between monthly and quarterly publication tables. The standard variance approximation of the GREG estimator assumes that these population totals are observed without error. This results in an underestimation of the variance of the GREG estimator. The variance approximation proposed in this paper results in more realistic standard errors for the quarterly GREG estimates.
    Release date: 2026-06-29

  • Articles and reports: 12-001-X202600100011
    Description: We construct a hybrid Bayesian method, which includes a differentially private mechanism, to mask Census county totals for a U.S. state on acreage of a commodity. We use surrogates for data collected at the farm level from a past U.S. Census of Agriculture to illustrate our procedure. We use two Bayesian small area models (parametric and mixture) to accommodate the smaller counties with fewer farms and some counties with large acres. In these models, the Laplace distribution provides a differentially private mechanism. In pre-processing, we also incorporate the Census weights to form the observed total acreage, a scaling factor to the Laplace mechanism for each county, a square-root transformation of the observed total acreage to avoid negative masked estimates especially for small counties, and the p-percent rule and the 3+ rule to partition the counties into suppressed counties, non-sensitive counties and sensitive counties. Because of difficulties in specifying and tuning the privacy budget (an unknown parameter), to balance security and utility, we specify a prior for the privacy budget, where the values are not specified, and the Gibbs sampler is used to fit the hierarchical Bayesian models. In post-processing, we use Bayesian predictive inference to obtain masked county acreages, and this includes a benchmarking so that the masked state total matches the observed state total. As a measure of reliability of the Bayesian procedure, we use the posterior coefficients of variation for the masked posterior means of the counties. As a measure of utility, we use the absolute relative errors for the individual counties, together with other global measures. For the sensitive counties, there are some differences between the two small area models but both are much better than an individual area model; the mixture model being the best compromise for security and utility.
    Release date: 2026-06-29

  • Articles and reports: 12-001-X202600100012
    Description: We propose small area estimators of general indicators in off-census years, which avoid the use of deprecated census microdata, but are nearly optimal in census years. The procedure is based on replacing the obsolete census file with a larger unit-level survey that adequately covers the areas of interest and contains the values of useful auxiliary variables. However, the minimal data requirement of the proposed method is a single survey with microdata on the target variable and suitable auxiliary variables for the period of interest. We also develop an estimator of the mean squared error (MSE) that accounts for the uncertainty introduced by the large survey used to replace the census of auxiliary information. Our empirical results indicate that the proposed predictors perform clearly better than the alternative predictors when census data are outdated, and are very close to optimal ones when census data are correct. They also illustrate that the proposed total MSE estimator corrects for the bias of purely model-based MSE estimators that do not account for the large survey uncertainty.
    Release date: 2026-06-29

  • Articles and reports: 12-001-X202600100013
    Description: In the age of big data, nonprobability surveys are becoming increasingly abundant. Data integration techniques involving both probability and nonprobability surveys are being extensively used for providing improved estimates for finite population estimation. While much of the existing research has focused on mitigating selection bias in nonprobability surveys, the issue of measurement error within these surveys remains relatively unexplored. Statistical methods devised with the purpose of reducing selection bias are appropriate for reliable estimation, only under the assumption of accuracy of survey responses. Motivated by a recent case study of Kennedy, Mercer and Lau (2024), our research addresses bias from both measurement and sampling errors in nonprobability surveys. In this article, we propose a new data integration method that uses multiple probability and nonprobability surveys and leverages machine learning models to construct a composite estimator. The proposed composite estimator integrates probability and nonprobability surveys, when both contain response variables of interest. We analyze the performance of this estimator in comparison to an existing composite estimator in literature, analytically as well as empirically, using multiple survey data from Kennedy et al. (2024). Finally, we identify conditions under which the proposed estimator outperforms estimators based solely on probability surveys.
    Release date: 2026-06-29

  • Journals and periodicals: 12-001-X
    Geography: Canada
    Description: The journal publishes articles dealing with various aspects of statistical development relevant to a statistical agency, such as design issues in the context of practical constraints, use of different data sources and collection techniques, total survey error, survey evaluation, research in survey methodology, time series analysis, seasonal adjustment, demographic studies, data integration, estimation and data analysis methods, and general survey systems development. The emphasis is placed on the development and evaluation of specific methodologies as applied to data collection or the data themselves.
    Release date: 2026-06-29

  • Table: 27-10-0005-01
    Geography: Canada
    Frequency: Annual
    Description: Federal expenditures on science and technology and its components in current dollars and 2012 constant dollars by type of expenditures for Canada from 1995 to today.
    Release date: 2026-06-29
Data (13,366)

Data (13,366) (12,370 to 12,380 of 13,366 results)

  • Table: 16-10-0070-01
    Frequency: Quarterly
    Description: This table contains 11 series, with data for years 1946 - 1994 (not all combinations necessarily have data for all years), and is no longer being released. This table contains data described by the following dimensions (Not all combinations are available): Geography (1 item: Canada); Production and stocks (11 items: Cereals ready to serve; Macaroni, et cetera, dry; Prepared cake mixes; Mixes, all other flour based; ...).
    Release date: 2000-02-19

  • Table: 16-10-0093-01
    Frequency: Quarterly
    Description: This table contains 7 series, with data for years 1946 - 1989 (not all combinations necessarily have data for all years), and is no longer being released. This table contains data described by the following dimensions (Not all combinations are available): Geography (1 item: Canada); Standard Classification of Goods (SCG) (7 items: Biscuits, soda; Biscuits, plain and fancy; Chewing gum; Cocoa powder; ...).
    Release date: 2000-02-19

  • Table: 22-10-0092-01
    Frequency: Monthly
    Description: Telephone statistics, by telephone operations for Canada from 1971 to 1998. (Terminated)
    Release date: 2000-02-19

  • Table: 12-10-0079-01
    Geography: Canada
    Frequency: Monthly
    Description:

    This table contains 8 series, with data for years 1946 - 1975 (not all combinations necessarily have data for all years). This table contains data described by the following dimensions (Not all combinations are available): Geography (1 item: Canada); Type of index (2 items: Index of prices; Index of physical volume); Index base period (2 items: 1948=100; 1968=100); Imports/exports (2 items: Imports; Exports).

    Release date: 2000-02-19

  • Table: 32-10-0278-01
    Frequency: Annual
    Description: This table contains 493 series, with data for years 1926 - 1972 (not all combinations necessarily have data for all years), and is no longer being released. This table contains data described by the following dimensions (Not all combinations are available): Geography (10 items: Canada;Prince Edward Island;Nova Scotia;New Brunswick; ...); Expenses and rebates (51 items: Total operating expenses;Taxes on owned farm land;Gross farm rent;Cash rent; ...).
    Release date: 2000-02-19

  • 12,376. Farm wages Archived
    Table: 32-10-0281-01
    Frequency: Quarterly
    Description: This table contains 48 series, with data for years 1940 - 1986 (not all combinations necessarily have data for all years), and is no longer being released. This table contains data described by the following dimensions (Not all combinations are available): Geography (8 items: Canada;Maritime provinces;Quebec;Ontario; ...); Farm wage rates (3 items: Monthly farm wages;Hourly farm wages;Daily farm wages); Components (2 items: With board;Without board).
    Release date: 2000-02-19

  • Table: 23-10-0093-01
    Geography: Canada
    Frequency: Monthly
    Description:

    This table contains 9 series, with data for years 1946 - 1995 (not all combinations necessarily have data for all years), and was last released on 2000-10-20. This table contains data described by the following dimensions (Not all combinations are available): Geography (1 item: Canada), Operating statistics (9 items: Railway revenues; Railway operating expenses; Railway operating income;Railway operating revenue; passenger; ...).

    Release date: 2000-02-19

  • Table: 23-10-0203-01
    Geography: Canada
    Frequency: Monthly
    Description:

    This table contains 16 series, with data for years 1975 - 1994 (not all combinations necessarily have data for all years), and was last released on 2000-05-25. This table contains data described by the following dimensions (Not all combinations are available): Geography (1 items: Canada), Operating statistics (16 items: Carriers reporting; Revenue earned, all services; Revenue earned, intercity and rural services; Revenue earned, urban-suburban services; ...).

    Release date: 2000-02-19

  • Table: 36-10-0384-01
    Geography: Canada
    Frequency: Annual
    Description: This table contains 159 series, with data for years 1946 - 1971 (not all combinations necessarily have data for all years). This table contains data described by the following dimensions (Not all combinations are available): Geography (1 items: Canada ...) Industry (159 items: Real domestic product; Agriculture; Forestry; Real domestic product less agriculture ...).
    Release date: 2000-02-19

  • Table: 36-10-0386-01
    Geography: Canada
    Frequency: Annual
    Description: This table contains 189 series, with data for years 1919 - 1971 (not all combinations necessarily have data for all years). This table contains data described by the following dimensions (Not all combinations are available): Geography (1 items: Canada ...) Industry (189 items: Index of industrial production; Mines (including milling); quarries and oil wells; Metal mines; Placer gold and gold quartz mines ...).
    Release date: 2000-02-19
Analysis (10,894)

Analysis (10,894) (10,690 to 10,700 of 10,894 results)

  • Articles and reports: 12-001-X198800214586
    Description:

    A generalized implementation of a method for performing automated coding is described. Traditionally, coding has been performed manually by specially trained personnel, but recently computerized systems have appeared which either eliminate or substantially reduce the need for manual coding. Typically, such systems are limited in use to those applications for which they were originally designed. The system presented here may be used by any application to perform coding of English or French text using any classification scheme.

    Release date: 1988-12-15

  • Articles and reports: 12-001-X198800214587
    Description:

    The QUID system, which was designed and developed by INSEE (Paris) Institut National de la Statistique et des Études Économiques- National Statistics and Economic Studies Institute, is an automatic coding system for survey data collected in the form of literal headings expressed in the terminology of the respondent. The system hinges on the use of a very wide knowledge base made up of real phrases coded by experts. This study deals primarily with the preliminary automatic standardization processing of the phrases, and then with the algorithm used to organize the phrase base into an optimized tree pattern. A sorting example is provided in the form of an illustration. At present, the processing of additional coding variables used to complement the information contained in the phrases presents certain difficulties, and these will be examined in detail. The QUID 2 project, an updated version of the system, will be discussed briefly.

    Release date: 1988-12-15

  • Articles and reports: 12-001-X198800214588
    Description:

    Suppose that undercount rates in a census have been estimated and that block-level estimates of the undercount have been computed. It may then be desirable to create a new roster of households incorporating the estimated omissions. It is proposed here that such a roster be created by weighting the enumerated households. The household weights are constrained by linear equations representing the desired total counts of persons in each estimation class and the desired total count of households. Weights are then calculated that satisfy the constraints while making the fitted table as close as possible to the raw data. The procedure may be regarded as an extension of the standard “raking” methodology to situations where the constraints do not refer to the margins of a contingency table. Continuous as well as discrete covariates may be used in the adjustment, and it is possible to check directly whether the constraints can be satisfied. Methods are proposed for the use of weighted data for various Census purposes, and for adjustment of covariate information on characteristics of omitted households, such as income, that are not directly considered in undercount estimation.

    Release date: 1988-12-15

  • Articles and reports: 12-001-X198800214589
    Description:

    The U.S. Bureau of the Census uses dual system estimates (DSEs) for measuring census coverage error. The dual system estimate uses data from the original enumeration and a Post Enumeration Survey. In measuring the accuracy of the DSE, it is important to know that the DSE is subject to several components of nonsampling error, as well as sampling error. This paper gives models of the total error and the components of error in the dual system estimates. The models relate observed indicators of data quality, such as a matching error rate, to the first two moments of the components of error. The propagation of error in the DSE is studied and its bias and variance are assessed. The methodology is applied to the 1986 Census of Central Los Angeles County in the Census Bureau’s Test of Adjustment Related Operations. The methodology also will be useful to assess error in the DSE for the 1990 census as well as other applications.

    Release date: 1988-12-15

  • Articles and reports: 12-001-X198800214590
    Description:

    This paper presents results from a study of the causes of census undercount for a hard-to-enumerate, largely Hispanic urban area. A framework for organizing the causes of undercount is offered, and various hypotheses about these causes are tested. The approach is distinctive for its attempt to quantify the sources of undercount and isolate problems of unique importance by controlling for other problems statistically.

    Release date: 1988-12-15

  • Articles and reports: 12-001-X198800214591
    Description:

    To estimate census undercount, a post-enumeration survey (PES) is taken, and an attempt is made to find a matching census record for each individual in the PES; the rate of successful matching provides an estimate of census coverage. Undercount estimation is performed within poststrata defined by geographic, demographic, and housing characteristics, X. Portions of X are missing for some individuals due to survey nonresponse; moreover, a match status Y cannot be determined for all individuals. A procedure is needed for imputing the missing values of X and Y. This paper reviews the imputation methods used in the 1986 Test of Adjustment Related Operations (Schenker 1988) and proposes two alternative model-based methods: (1) a maximum-likelihood contingency-table estimation procedure that ignores the missing-data mechanism; and (2) a new Bayesian contingency table estimation procedure that does not ignore the missing-data mechanism. The first method is computationally simpler, but the second is preferred on conceptual and scientific grounds.

    Release date: 1988-12-15

  • Articles and reports: 12-001-X198800214592
    Description:

    There are persuasive arguments for and against adjustment of the U.S. decennial census counts, although many of them are based on political rather than technical considerations. The decision whether or not to adjust depends crucially on the method of adjustment. Moreover, should adjustment take place using say a synthetic-based or a regression-based method, at which level should this occur and how should aggregation and disaggregation proceed? In order to answer these questions sensibly, a model of undercount errors is needed which is “level-consistent” in the sense that it is preserved for areas at the national, state, county, etc. level. Such a model is proposed in this article; like subareas are identified with strata such that within a stratum the subareas’ adjustment factors have a common stratum mean and have variances inversely proportional to their census counts. By taking into account sampling of the areas (e.g., by dual-system estimation), empirical Bayes estimators that combine information from the stratum average and the sample value, can be constructed. These estimators are evaluated at the state level (51 states, including Washington, D.C.), and stratified on race/ethnicity (3 strata) using data from the 1980 postenumeration survey (PEP 3-8, for the noninstitutional population).

    Release date: 1988-12-15

  • Articles and reports: 12-001-X198800214593
    Description:

    In Australia, population estimates have been obtained from census counts, incorporating an adjustment for under-enumeration in 1976, 1981 and 1986. The adjustments are based on the results of a Post Enumeration Survey and demographic analysis. This paper describes the methods used and the results obtained in adjusting the 1986 census. The formal use of sex ratios as suggested by Wolter (1986) is examined as a possible improvement of the less formal use made of these ratios in adjusting census counts.

    Release date: 1988-12-15

  • Articles and reports: 12-001-X198800214594
    Description:

    A significant increase in coverage error in the 1986 Census is revealed by both the Reverse Record Check and the demographic method presented in this paper. Considerable attention is paid to an evaluation of the various components of population growth, especially interprovincial migration. The paper concludes with an overview of two alternative methods for generating postcensal estimates: the currently-in-use, census-based model, and a flexible model using all relevant data in combination with the census.

    Release date: 1988-12-15

  • Articles and reports: 12-001-X198800214595
    Description:

    Estimates of undercoverage in the Canadian Census of Population have been produced for each Census since 1961, using a Reverse Record Check method. The reliability of the estimates is important to how they are used to assess the quality of the Census data and to identify significant causes of coverage error. It is also critical to the development of methods and procedures to improve coverage for future Censuses. The purpose of this paper is to identify potential sources of error in the Reverse Record Check, which should be understood and addressed, where possible, in using this method to estimate coverage error.

    Release date: 1988-12-15
Reference (2,038)

Reference (2,038) (970 to 980 of 2,038 results)

  • Surveys and statistical programs – Documentation: 11-522-X19990015658
    Description:

    Radon, a naturally occurring gas found at some level in most homes, is an established risk factor for human lung cancer. The U.S. National Research Council (1999) has recently completed a comprehensive evaluation of the health risks of residential exposure to radon, and developed models for projecting radon lung cancer risks in the general population. This analysis suggests that radon may play a role in the etiology of 10-15% of all lung cancer cases in the United States, although these estimates are subject to considerable uncertainty. In this article, we present a partial analysis of uncertainty and variability in estimates of lung cancer risk due to residential exposure to radon in the United States using a general framework for the analysis of uncertainty and variability that we have developed previously. Specifically, we focus on estimates of the age-specific excess relative risk (ERR) and lifetime relative risk (LRR), both of which vary substantially among individuals.

    Release date: 2000-03-02

  • Surveys and statistical programs – Documentation: 11-522-X19990015660
    Description:

    There are many different situations in which one or more files need to be linked. With one file the purpose of the linkage would be to locate duplicates within the file. When there are two files, the linkage is done to identify the units that are the same on both files and thus create matched pairs. Often records that need to be linked do not have a unique identifier. Hierarchical record linkage, probabilistic record linkage and statistical matching are three methods that can be used when there is no unique identifier on the files that need to be linked. We describe the major differences between the methods. We consider how to choose variables to link, how to prepare files for linkage and how the links are identified. As well, we review tips and tricks used when linking files. Two examples, the probabilistic record linkage used in the reverse record check and the hierarchical record linkage of the Business Number (BN) master file to the Statistical Universe File (SUF) of unincorporated tax filers (T1) will be illustrated.

    Release date: 2000-03-02

  • Surveys and statistical programs – Documentation: 11-522-X19990015662
    Description:

    As the availability of both health utilization and outcome information becomes increasingly important to health care researchers and policy makers, the ability to link person-specific health data becomes a critical objective. This type of linkage of population-based administrative health databases has been realized in British Columbia. The database was created by constructing an historical file of all persons registered with the health care system, and then by probabilistically linking various program files to this 'coordinating' file. The first phase of development included the linkage of hospital discharge data, physician billing data, continuing care data, data about drug costs for the elderly, births data and deaths data. The second phase of development has seen the addition data sources external to the Ministry of Health including cancer incidence data, workers' compensation data, and income assistance data.

    Release date: 2000-03-02

  • Surveys and statistical programs – Documentation: 11-522-X19990015664
    Description:

    Much work on probabilistic methods of linkage can be found in the statistical literature. However, although many groups undoubtedly still use deterministic procedures, not much literature is available on these strategies. Furthermore there appears to exist no documentation on the comparison of results for the two strategies. Such a comparison is pertinent in the situation where we have only non-unique identifiers like names, sex, race etc. as common identifiers on which the databases are to be linked. In this work we compare a stepwise deterministic linkage strategy with the probabilistic strategy, as implemented in AUTOMATCH, for such a situation. The comparison was carried out on a linkage between medical records from the Regional Perinatal Intensive Care Centers database and education records from the Florida Department of Education. Social security numbers, available in both databases, were used to decide the true status of the record pair after matching. Match rates and error rates for the two strategies are compared and a discussion of their similarities and differences, strengths and weaknesses is presented.

    Release date: 2000-03-02

  • Surveys and statistical programs – Documentation: 11-522-X19990015666
    Description:

    The fusion sample obtained by a statistical matching process can be considered a sample out of an artificial population. The distribution of this artificial population is derived. If the correlation between specific variables is the only focus the strong demand for conditional independence can be weakened. In a simulation study the effects of violations of some assumptions leading to the distribution of the artificial population are examined. Finally some ideas concerning the establishing of the claimed conditional independence by latent class analysis are presented.

    Release date: 2000-03-02

  • Surveys and statistical programs – Documentation: 11-522-X19990015668
    Description:

    Following the problems with estimating underenumeration in the 1991 Census of England and Wales the aim for the 2001 Census is to create a database that is fully adjusted to net underenumeration. To achieve this, the paper investigates weighted donor imputation methodology that utilises information from both the census and census coverage survey (CCS). The US Census Bureau has considered a similar approach for their 2000 Census (see Isaki et al 1998). The proposed procedure distinguishes between individuals who are not counted by the census because their household is missed and those who are missed in counted households. Census data is linked to data from the CCS. Multinomial logistic regression is used to estimate the probabilities that households are missed by the census and the probabilities that individuals are missed in counted households. Household and individual coverage weights are constructed from the estimated probabilities and these feed into the donor imputation procedure.

    Release date: 2000-03-02

  • Surveys and statistical programs – Documentation: 11-522-X19990015670
    Description:

    To reach their target audience efficiently, advertisers and media planners need information on which media their customers use. For instance, they may need to know what percentage of Diet Coke drinkers watch Baywatch, or how many AT&T customers have seen an advertisement for Sprint during the last week. All the relevant data could theoretically be collected from each respondent. However, obtaining full detailed and accurate information would be very expensive. It would also impose a heavy respondent burden under current data collection technology. This information is currently collected through separate surveys in New Zealand and in many other countries. Exposure to the major media is measured continuously, and product usage studies are common. Statistical matching techniques provide a way of combining these separate information sources. The New Zealand television ratings database was combined with a syndicated survey of print readership and product usage, using statistical matching. The resulting Panorama service meets the targeting information needs of advertisers and media planners. It has since been duplicated in Australia. This paper discusses the development of the statistical matching framework for combining these databases, and the heuristics and techniques used. These included an experiment conducted using a screening design to identify important matching variables. Studies evaluating and validating the combined results are also summarized. The following three major evaluation criteria were used; accuracy of combined results, statibility of combined results and the preservation of currency results from the component databases. The paper then discusses how the prerequisites for combining the databases were met. The biggest hurdle at this stage was the differences between the analysis techniques used on the two component databases. Finally, suggestions for developing similar statistical matching systems elsewhere will be given.

    Release date: 2000-03-02

  • Surveys and statistical programs – Documentation: 11-522-X19990015672
    Description:

    Data fusion as discussed here means to create a set of data on not jointly observed variables from two different sources. Suppose for instance that observations are available for (X,Z) on a set of individuals and for (Y,Z) on a different set of individuals. Each of X, Y and Z may be a vector variable. The main purpose is to gain insight into the joint distribution of (X,Y) using Z as a so-called matching variable. At first however, it is attempted to recover as much information as possible on the joint distribution of (X,Y,Z) from the distinct sets of data. Such fusions can only be done at the cost of implementing some distributional properties for the fused data. These are conditional independencies given the matching variables. Fused data are typically discussed from the point of view of how appropriate this underlying assumption is. Here we give a different perspective. We formulate the problem as follows: how can distributions be estimated in situations when only observations from certain marginal distributions are available. It can be solved by applying the maximum entropy criterium. We show in particular that data created by fusing different sources can be interpreted as a special case of this situation. Thus, we derive the needed assumption of conditional independence as a consequence of the type of data available.

    Release date: 2000-03-02

  • Surveys and statistical programs – Documentation: 11-522-X19990015674
    Description:

    The effect of the environment on health is of increasing concern, in particular the effects of the release of industrial pollutants into the air, the ground and into water. An assessment of the risks to public health of any particular pollution source is often made using the routine health, demographic and environmental data collected by government agencies. These datasets have important differences in sampling geography and in sampling epochs which affect the epidemiological analyses which draw them together. In the UK, health events are recorded for individuals, giving cause codes, a data of diagnosis or death, and using the unit postcode as a geographical reference. In contrast, small area demographic data are recorded only at the decennial census, and released as area level data in areas distinct from postcode geography. Environmental exposure data may be available at yet another resolution, depending on the type of exposure and the source of the measurements.

    Release date: 2000-03-02

  • Surveys and statistical programs – Documentation: 11-522-X19990015676
    Description:

    As the population ages, a greater demand for long-term care services and, in particular, nursing homes is expected. Policy analysts continue to search for alternative, less costly forms of care for the elderly and have attempted to develop programs to delay or prevent nursing-home entry. Health care administrators required information for planning the future demand for nursing-home services. This study assesses the relative importance of predisposing, enabling, and need characteristics in predicting and understanding nursing-home entry.

    Release date: 2000-03-02