All time references are in CEST
Advances in Life History Data Collection |
|
| Session Organiser | Dr Sebastian Lang (Leibniz Institute for Educational Trajectories (LIfBi)) |
| Time | Tuesday 18 July, 09:00 - 10:30 |
| Room | |
The collection of life history data presents unique methodological challenges within survey research, particularly when conducted through self-administered modes. Life history data, which capture the sequence, timing, duration, and likely other detailed information about events throughout an individual's life, are essential for understanding complex social phenomena such as education and employment trajectories. However, collecting accurate and reliable data on past life events is notoriously difficult. Respondents often face significant cognitive burdens when asked to recall events that may have occurred years or even decades ago. Factors such as memory lapses, recall bias, and inconsistencies, such as in the order of events, can all contribute to inaccuracies that can affect the quality and reliability of the data collected.
Retrospective surveys using traditional questionnaire methods are particularly susceptible to these problems. Various methods, such as life history calendars (LHCs), have been proposed to mitigate some of these challenges. At the same time, in an era of declining response rates and rising survey costs in interviewer-administered modes, self-administered modes, and consequently web-based surveys, are on the rise. However, the implementation of these methods in self-administered surveys presents new challenges. Without interviewer assistance, respondents may struggle to accurately report the information they are asked to provide, potentially leading to higher error rates and incomplete data.
In addition, the lack of standardization of these applications across studies makes it difficult to comprehensively assess their effectiveness. Addressing these challenges, including improving respondent recall and ensuring data consistency and quality, especially in self-administered modes, remains critical to advancing life history data collection. This session will explore recent innovations and ongoing challenges in life history data collection, with a focus on self-administered survey modes.
Keywords: life history data, data quality, web-based surveys, self-administered surveys
Dr Johann Carstensen (German Centre for Higher Education Research and Science Studies (DZHW)) - Presenting Author
Dr Sebastian Lang (Leibniz Institute for Educational Trajectories (LIfBi))
Mr Christian Friedrich (Federal Office of Consumer Protection and Food Safety)
Mr Christian Meisner (German Centre for Higher Education Research and Science Studies)
Ms Andrea Schulze (District administration of the Wetterau district)
Many data collection efforts in the social sciences involve the collection of some kind of life history data. Since life courses are complex and reporting on them carries a heavy response burden, methodological requirements are especially high in this case. For self-administered surveys, there is a noticeable lack of appropriate methods to collect this kind of data.
The contribution describes the development process and its results in the endeavor to design a novel survey instrument for the collection of life course data in self-administered online surveys. We therefore outline the theoretical and methodological desiderata on which the development of our life course module was based. We identify requirements and technical restrictions that had to be overcome in order to design a module that fulfills all of the requirements. The solution to a number of central restrictions in our survey software includes the usage of JSON objects as content of string variables. They enable us to make modifications such as splitting episodes on the fly and presenting the respondents with the modified survey data.
The finished module includes a Life History Calendar (LHC) that is usable both to enter life course data and to present the modified data. However, while LHCs can only provide very little information about an episode (such as start and end date), a combination with a question-list can enable researchers to assign characteristics to episodes, such as income and employment satisfaction as characteristics of an employment episode. This combination is especially beneficial to reduce adverse response effects while keeping up the amount of information requested from respondents. In our newly developed life history survey instrument, LHC and additional question-lists can be combined and information and structure gathered in one of those elements can be displayed dynamically in the other.
Ms danielle remmerswaal (Utrecht University) - Presenting Author
Professor Bella Struminskaya (Utrecht University)
Professor Barry Schouten (Statistics Netherlands)
Traditionally, diary studies are used to capture time use behaviour, such as traveling, physical activity, and budgeting, that is otherwise prone to recall errors, telescoping and other measurement errors when asked retrospectively. However, the traditional diary modes, web and paper, pose a high burden for respondents, resulting in nonresponse and measurement errors such as rounding, and underreporting. Given these limitations and the technological advances of smartphones and other smart devices, growing attention is given to surveys on smart devices (i.e. smart surveys). Smart surveys can use sensors on devices, and the device’s intelligence to store, compute, and predict data.
Two example applications in official statistics are a travel diary and a budget diary in which the manual entry of data is (partially) replaced by sensors on a (smart) device. Smartphone travel apps can make use of geolocation sensors present on the smartphone to construct a travel diary. In a budget app respondents can use their smartphone camera to scan receipts, which can then be processed into expenses with a combination of Optical Character Recognition (OCR) technology and machine learning.
The biggest advantages is the lower response burden for entering data. Earlier attempts to reduce the burden of traditional diary studies have mainly focused on minimizing the number of diary days that respondents have to report on. Despite ample evidence that a short reporting period is not representative of respondent's average behaviour. To investigate the desirable number of diary days from a statistical perspective, we calculate the reliability of multiple-day measures by separating the within-person and between-person variances for statistical estimates of interest.
We aim to provide guidelines for researchers on selecting the appropriate duration for (smart) diary studies, taking into consideration the participant burden, the characteristics of the diary instrument, and analytical objectives.
Dr Alessandra Gaia (CLS, UCL)
Mr Matt Brown (CLS, UCL)
Dr Michaela Sedovic (CLS, UCL) - Presenting Author
Professor George Ploubidis (CLS, UCL)
Professor Alissa Goodman (CLS, UCL)
Dr Darina Peycheva (UCL Institute of Epidemiology & Health)
Childhood circumstances have a significant impact on adult life and for this reason, many studies of older adults collect information about the early years of life. For example, the English Longitudinal Study of Ageing (ELSA), the U.S. Health and Retirement Study (HRS) and the Survey of Health, Ageing and Retirement in Europe (SHARE), have all administered life history questionnaires to collect information about life experiences prior to joining the study. However, little is known about the accuracy with which adults can recall information about their childhood, and whether some aspects of childhood can be more accurately recalled than others.
The National Child Development Study (NCDS) has been following the lives of over 17,000 people in Britain born in a single week of 1958. The tenth (Age 62) follow-up of the study includes a retrospective questionnaire about childhood circumstances. The same questions were answered contemporaneously by participants themselves, parents or teachers during childhood follow-ups at ages 7, 11 and 16. The retrospective questionnaire covers many domains typically included in life history questionnaires, such as housing, parental employment, and health, but also topics not typically explored because of uncertainty regarding the accuracy with which they can be recalled, such as measures of behaviour and psychological distress.
This research compares retrospective responses with data gathered prospectively, providing a unique opportunity to explore the accuracy with which a range of childhood circumstances and experiences can be recalled.
Additionally, we investigate whether using prospectively or retrospectively collected data results in substantive differences in the estimated associations between childhood circumstances and later life outcomes such as health and income.
We will conclude by discussing the implications of our results for survey practice, particularly in the design of studies on the old age population, and for data users analysing retrospectively collected data.
Dr Sebastian Lang (Leibniz Institute for Educational Trajectories)
Dr Heike Spangenberg (German Centre for Higher Education Research and Science Studies)
Dr David Ohlendorf (German Centre for Higher Education Research and Science Studies)
Mr Heiko Quast (German Centre for Higher Education Research and Science Studies)
Ms Leena Maaß (German Centre for Higher Education Research and Science Studies) - Presenting Author
The collection of life history data is a key instrument in social sciences, but it poses challenges for researchers and respondents, especially as it is usually done retrospectively. In particular, the processes in autobiographical memory associated with the collection of life history data are prone to error and can impact data quality. To address these challenges, the Life History Calendar (LHC) has been established as a valuable tool for collecting life history data. Considering the costs associated with face-to-face surveys conducted by interviewers and the decline in participation rates, however, self-administered web surveys (CAWI) are becoming increasingly important.
Against this background, we will develop a web application that incorporates a dynamic life history survey module, including an LHC, explicitly tailored for self-administration. In contrast to traditional (panel-)studies, we expect several advantages: this application will allow respondents to report changes in their life course at any time and shortly after they occur. This should enable respondents to provide their answers in less time, enhance usability and alleviate the response burden for the main survey. Furthermore, we expect the web application to simplify panel maintenance in terms of contact data maintenance, which should have a positive effect on the deliverability of survey materials.
To test these expectations, we will design two experiments in which respondents will be randomly divided into three groups: two treatment groups, collecting life history data prospectively, and one control group, collecting the same data retrospectively.
In our contribution, we will provide initial insights into the survey tool and the web application. Moreover, we will present the experimental design in detail and elaborate on our planned research on the effect on response behavior and rates, data quality and response burden.
Dr Cristian Domarchi (University of Southampton) - Presenting Author
Dr Olga Maslovskaya (University of Southampton)
Professor Lisa Calderwood (Centre for Longitudinal Studies, University College London)
Mr Matt Brown (Centre for Longitudinal Studies, University College London)
The study of life events has long been a central topic in social science. While longitudinal studies gather information about life events prospectively, most research relies on retrospective data collection, which depends on respondents’ ability to accurately recall past experiences. This reliance can lead to misreporting or omissions due to recall bias.
Surveys collect retrospective data using various methods. Traditional question-based approaches involve sequences of questions about events in different life domains (e.g., housing, relationships, employment). However, these methods may struggle to capture complex interrelated events over extended periods. An alternative is the calendar-based approach, which uses graphical representations of the reference period and life domains to facilitate memory recall through parallel cueing.
Calendar-based approaches, such as Life History Calendars (LHCs) and Event History Calendars (EHCs), enhance data quality by providing visual temporal cues and promoting the retrieval of autobiographical memories. These methods have predominantly been used in interviewer-administered surveys, where interaction aids recall. However, the rise of web-based surveys has increased interest in adapting these techniques for self-administered online contexts, despite the challenges involved.
In this presentation, we present findings from a recent evidence review conducted as part of the Survey Futures project, which gathered evidence from 13 studies that have collected retrospective life history data online. The studies included high-quality probability-based longitudinal and cross-sectional surveys, as well as surveys with experimental designs. We will summarise the different approaches employed in these studies and present practical recommendations based on their findings, focusing on the key design features, which were found to improve recall and data quality.