Project Introduction

Social survey related Project

Social Survey Historical Record Archive

Project Summary

Regarding various social surveys conducted in the form of joint research, in addition to publishing microdata, the Center for Social Data Structuring (CSDS) and the Institute of Statistical Mathematics (ISM) provide users with various supplementary information (survey ancillary information) that contributes to detailed analysis and discussion of survey results for collaboration. This is intended to be used as a reference for historical research on social surveys and future social survey design.

Project Background and Purpose: The Importance of Survey Ancillary Information

The primary outcome of a social survey is its results, and the manner in which the survey is tabulated and reported, as well as its data analysis, determine the value of such results. However, although the process for reaching such results is partially recorded in the report, the published information is insufficient most times.
Let us consider the survey preparation process regarding questionnaire preparation (measurement process) and subject selection (sampling) separately. Although it varies according to the subject and purpose of the survey, the question of how the items are created (taken from past surveys or newly written) is important. For an international comparative survey, a crucial element in determining the quality of the survey is the process of preparing the questionnaire in multiple languages, including translation and back-translation. Meanwhile, various considerations such as the concept on which the sample design was based, e.g., the selection of units for multi-stage sampling, the criteria for layers and their historical records, the frame concept and register access, etc., are included during sampling.
When it comes to individual social survey results, viewing historical records that thoroughly explain the preparing process may not be possible. In this project, we emphasize the importance of information ancillary for (the implementation of) this kind of survey. The main purpose is to preserve such information so that it can be accessed as a reference for future survey design, i.e., to archive the ancillary survey information and make it publicly available for collaboration.
Since its inception, the Institute of Statistical Mathematics (ISM) has used many historical records from the past to develop various survey products. This project inherits many of these records, and thus, rather than social survey “materials,” it refers to “historical records” in its name. The nature of these records, however, is not necessarily historical.

Anticipated Results and Outcomes (examples of survey ancillary information for publication and collaboration)

Various Types of Survey Report

As part of this project, various research reports that had not been fully published by other means have been made public.
For example:

Ancillary Documents at the Time of Survey Planning

Various documents were retained at the time of survey planning although they may not be present in the survey reports. These documents may also provide important insights for the future survey researchers .

Survey Planning Documents

There may be historical records that describe the entire survey plan. Nowadays, before conducting a survey, survey plan documents should be submitted for “ethics review.” However, these documents will not necessarily be described in the report.

• Example: 1948 Literacy Survey for Japanese People Planning Document

  Recently, a survey planning document for the Literacy Survey for Japanese People was discovered in the Research Materials Room of the National Institute for Japanese Language and Linguistics (NINJAL). This survey, conducted in 1948, is famous in the history of social surveys. These archival records were published by the National Institute for Japanese Language and Linguistics (NINJAL) with the cooperation of the Center for Social Data Structuring (CSDS).
  (See “Literacy Research Program”)

Foundational Historical Records Considering Sample Design in Each Type of Survey (under preparation)

When designing a survey sample, “layering” and “multi-stage sampling” are commonly used techniques. This kind of sample design generally involves gathering a several information on the sample target, particularly regional characteristics, based on official statistics. By preserving historical records describing these materials, the policies and other details that were prevalent at the time of sample design can be known. This, in turn, can be used as a reference when designing samples for similar surveys thereafter.

List of Survey Sites (under preparation)

Social surveys aimed at individuals collect responses in a hierarchical structure, such as by “survey site – individual,” through a process of two-stage sampling. Frequently, however, this survey site information does not include microdata. Information about the survey site provides important variables when examining the data properties and performing a statistical analysis. However, disclosing the details of site information increases the risk of individuals being identified from the micro-data. It is necessary, therefore, to be careful when disclosing information.

Historical Records Prepared after the Survey Was Conducted (under preparation)

In the case of survey items with an open-ended answer format, it is common for survey administrators to “code” responses after the fact for tabulation purposes. While the open-ended answer content is recorded in the survey report in some cases, historically speaking, the process has involved transcribing answers onto cards or other documents and then organizing them appropriately. Historical records organized in such a manner could be made available for collaboration.