There are various ways to approach respondents for a questionnaire-based survey: respondents can complete a questionnaire with the help from an interviewer, but they can also do it themselves. Interviewers can be involved in two ways: face-to-face or via the telephone. There are also two ways for respondents to complete the questionnaire themselves: on paper, or online. This paper describes the applications, advantages and disadvantages of all these possibilities.
Roughly speaking, a survey consists of a design, an approach strategy and a questionnaire. In this report, the authors describe the development of a questionnaire in detail. They explain the various steps in the development process, the design of the questions, formulation of answer options, explanations and instructions, and the design of the questionnaire both on paper and for the internet. The final part of the report examines various testing methods.
If no other sources are available to measure certain variables of a target population, Statistics Netherlands often uses a random sample, i.e. a sample of units from the population drawn in accordance with a well-prescribed selection mechanism. This report describes the various sample designs and weighting methods regularly used at Statistics Netherlands.
Statistics Netherlands aims to make as much use as possible of data sources that have been collected by others (i.e. not by Statistics Netherlands). Secondary data collection is the process of collecting such sources. This document discusses the advantages and disadvantages of this approach.
Survey researchers aim for the highest possible response rate, given the budget of their survey. This report explains why high response rates are important and presents overviews of measures to increase response rates for CAPI, CATI and web surveys. The overviews describe both general measures and specific measures taken by Statistics Netherlands.
The primary goal of business panels is to obtain estimates of the turnover, including estimates of the changes of the level of the turnover. Panels can also be used for obtaining observations on prices of goods or services for the producer price index (PPI) or the price index for services (SPI). The present document pays special attention to the variance of the estimator of a change of the level of the study variable in a given period. Complicating factors in a panel as panel attrition; births and deaths; and the fact that enterprises may migrate between strata are taken into account. The consequences of combining or splitting up enterprises are discussed as well. Due to the specific variance analysis of the estimates thus obtained, the document has a rather technical character.
Statistics Netherlands aims to produce high quality statistics. To do this it collects a vast amount of data on persons, households and companies. It is important that interviewers used to collect this information receive appropriate training. Comprehensive interviewer training increases response and is important for the quality of the information collected. This report examines the important aspects of the design and content of interviewer training courses. It also looks at how Statistics Netherlands trains its interviewers.
Survey researchers are confronted by a number of phenomena which may affect the quality of the results of their surveys. These phenomena may lead to a greater margin of error of the results, and are therefore called survey errors. Some errors can be prevented by taking them into account in the survey design and implementation. However, some errors cannot be prevented, and then their effects have to be minimised. This report gives a systematic overview of the various sources of errors.
Statistics Netherlands aims to produce high-quality and undisputed statistics. High response rates for its surveys are an important element in this respect. Unfortunately, respondents do not always automatically participate in surveys. By taking measures to enhance response, Statistics Netherlands can convince people to complete and return the questionnaire. This report explains a number of ways to increase response rates in business surveys.
XBRL (eXtensible Business Reporting Language) is an instrument used across government agencies to exchange financial data between these agencies and private companies. This report explains how Statistics Netherlands uses XBRL to collect data. It focuses on the construction of taxonomies and processing of instances.