Search

This paper studies identification and estimation of installed-base effects for product adoption using group-level panel data in the presence of endogenous sample attrition and homophily. After expl...

Background A major challenge of prospective cohort studies is attrition in follow-up surveys. This study investigated attrition in a prospective cohort comprised of medical graduates in China. We described status of attrition, identified participants with higher possibility of attrition, and examined if attrition affect the estimation of the key outcome measures. Methods The cohort study recruited 3,620 new medical graduates from four medical universities in central and western China between 2015 and 2019. Online follow-up surveys were conducted on an annual basis. Follow-up status was defined as complete (meaning that the participant completed all the follow-up surveys) and incomplete, while incomplete follow-up was further divided into ‘always-out’, ‘rejoin’ and ‘other’. Multivariable logistic and linear regressions were used to examine factors predicting attrition and the influence on the outcome measures of career development. Results 2364 (65.3%) participants completed all follow-up surveys. For those with incomplete follow-up, 520 (14.4%) were ‘always-out’, 276 (7.6%) rejoined in the 2020 survey. Willingness to participate in residency training (OR=0.80, 95%CI[0.66 - 0.98]) and willingness to provide sensitive information in the baseline survey predicted a lower rate of attrition (providing scores for university entrance exam OR=0.82, 95%CI[0.69 - 0.97]]; providing contact information (OR=0.46, 95%CI[0.32 - 0.66]); providing household income (OR=0.60, 95%CI[0.43 - 0.84]). Participants with compulsory rural service (OR=1.52, 95%CI[1.05 - 2.19]) and those providing university entrance scores (OR=1.64, 95%CI[1.15-2.33)) were more likely to rejoin in the follow-up survey. These factors associated with follow-up status did not have significant impact on key outcome measures of career development. Conclusions Graduates who were unwilling to participate in residency training or not providing sensitive information should be targeted early in the cohort study to reduce attrition. More information about the study should be provided to those graduates early to facilitate their understanding of the meaning in participation. On the contrary, medical graduates with compulsory rural service and those who provided university entrance scores were more likely to rejoin in the cohort. The research team should invest more effort in contacting those graduates and returned them to the cohort.

Sample attrition increases the risk of statistical bias and hinders the ability to plausibly estimate causal effects when patterns of nonresponse are correlated with key variables of interest. Draw...

Objectives Cognitive behavioural therapy is commonly used to treat chronic fatigue syndrome and has been shown to be effective for reducing fatigue and improving physical functioning. Most of the evidence on the effectiveness of cognitive behavioural therapy for chronic fatigue syndrome is from randomised control trials, but there are only a few studies in naturalistic treatment settings. Our aim was to examine the effectiveness of cognitive behavioural therapy for chronic fatigue syndrome in a naturalistic setting and examine what factors, if any, predicted outcome. Design Using linear mixed effects analysis, we analysed patients' self-reported symptomology over the course of treatment and at three-month follow-up. Furthermore, we explored what baseline factors were associated with improvement at follow-up. Setting Data were available for 995 patients receiving cognitive behavioural therapy for chronic fatigue syndrome at an outpatient clinic in the UK. Participants Participants were referred consecutively to a specialist unit for chronic fatigue or chronic fatigue syndrome. Main outcome measures Patients were assessed throughout their treatment using self-report measures including the Chalder Fatigue Scale, 36-item Short Form Health Survey, Hospital Anxiety and Depression Scale and Global Improvement and Satisfaction. Results Patients’ fatigue, physical functioning and social adjustment scores significantly improved over the duration of treatment with medium to large effect sizes (|d| = 0.45–0.91). Furthermore, 85% of patients self-reported that they felt an improvement in their fatigue at follow-up and 90% were satisfied with their treatment. None of the regression models convincingly predicted improvement in outcomes with the best model being (R2 = 0.137). Conclusions Patients’ fatigue, physical functioning and social adjustment all significantly improved following cognitive behavioural therapy for chronic fatigue syndrome in a naturalistic outpatient setting. These findings support the growing evidence from previous randomised control trials and suggest that cognitive behavioural therapy could be an effective treatment in routine treatment settings.

We show that survey and administrative data-based estimates of a panel data model of earnings, employment, and outmigration yield very different qualitative and quantitative predictions. Survey-based estimates substantially overpredict outmigration, in particular for lower performing immigrants. Consequently, employment and earnings of immigrants who remain in the country are overpredicted relative to model predictions from administrative data. Importantly, estimates from both data sources find opposite self-selection mechanisms into outmigration. Differences hold despite using the same cohort, survey period, and observable characteristics. Differences in predictions are driven by difficulties of properly separating non-random sample attrition from selective outmigration in survey data.

We present direct evidence on the link between children’s patience and educational-track choices years later. Combining an incentivized patience measure of 493 primary-school children with their high-school track choices taken at least three years later at the end of middle school, we find that patience significantly predicts choosing an academic track. This relationship remains robust after controlling for a rich set of covariates, such as family background, school-class fixed effects, risk preferences, and cognitive abilities, and is not driven by sample attrition. Accounting for middle-school GPA as a potential mediating factor suggests a direct link between patience and educational-track choice.

We show that survey and administrative data-based estimates of a panel data model of earnings, employment, and outmigration yield very different qualitative and quantitative predictions. Survey-based estimates substantially overpredict outmigration, in particular for lower performing immigrants. Consequently, employment and earnings of immigrants who remain in the country are overpredicted relative to model predictions from administrative data. Importantly, estimates from both data sources find opposite self-selection mechanisms into outmigration. Differences hold despite using the same cohort, survey period, and observable characteristics. Differences in predictions are driven by difficulties of properly separating non-random sample attrition from selective outmigration in survey data.

Adverse childhood experiences (ACEs) are associated with numerous physical and mental health issues in children and adults. The effect of ACEs on development of childhood obesity is less understood. This systematic review was undertaken to synthesize the quantitative research examining the relationship between ACEs and childhood obesity. PubMed, PsycInfo, and Web of Science were searched in July 2020; Rayyan was used to screen studies, and the Newcastle-Ottawa Scale was used to assess risk of bias. The search resulted in 6,966 studies screened at title/abstract and 168 at full-text level. Twenty-four studies met inclusion criteria. Study quality was moderate, with greatest risk of bias due to method of assessment of ACEs or sample attrition. Findings suggest ACEs are associated with childhood obesity. Girls may be more sensitive to obesity-related effects of ACEs than boys, sexual abuse appears to have a greater effect on childhood obesity than other ACEs, and co-occurrence of multiple ACEs may be associated with greater childhood obesity risk. Further, the effect of ACEs on development of childhood obesity may take 2-5 years to manifest. Considered collectively, findings suggest a need for greater attention to ACEs in the prevention and treatment of childhood obesity.

Studies have consistently found a longitudinal association between e-cigarette use (vaping) and cigarette smoking. Many have interpreted such association as causal. This systematic review and meta-analysis evaluated the plausibility of a causal interpretation by (1) estimating the effect of adolescent vaping on smoking initiation, adjusted for study quality characteristics, (2) evaluating the sufficiency of adjustment for confounding based on the social development model (SDM) and the social ecological model (SEM) and E-value analyses and (3) investigating sample attrition and publication bias.Systematic review and meta-analysis of longitudinal studies that examined the association between e-cigarette use at baseline and smoking at follow-up. Participants were non-smokers aged 18 at baseline.Meta-analysis of 11 studies showed a significant longitudinal association between vaping and smoking [adjusted odds ratio (aOR) = 2.93, 95% confidence interval (CI) = 2.22, 3.87]. Studies with sample sizes 1000 had a significantly higher odds ratio (OR = 6.68, 95% CI = 3.63, 12.31) than studies with sample sizes 1000 (OR = 2.49, 95% CI = 1.97, 3.15). Overall, the attrition rate was very high (median = 30%). All but one study reported results from complete sample analysis, despite those dropping out having higher risk profiles. Only two studies comprehensively adjusted for confounding. The median E-value was 2.90, indicating that the estimates were not robust against unmeasured confounding.There is a longitudinal association between adolescent vaping and smoking initiation; however, the evidence is limited by publication bias, high sample attrition and inadequate adjustment for potential confounders.

Missing data are hard to avoid, or even inevitable, in longitudinal learning diagnosis and other longitudinal studies. Sample attrition is one of the most common missing patterns in practice, which refers to students dropping out before the end of the study and not returning. This brief research aims to examine the impact of a common type of sample attrition, namely, individual-level random attrition, on longitudinal learning diagnosis through a simulation study. The results indicate that (1) the recovery of all model parameters decreases with the increase of attrition rate; (2) comparatively speaking, the attrition rate has the greatest influence on diagnostic accuracy, and the least influence on general ability; and (3) a sufficient number of items is one of the necessary conditions to counteract the negative impact of sample attrition.

International audience; BACKGROUND: Findings from efficacy trials of group psychoeducation (PE) for bipolar disorders (BD) led to its inclusion in evidence-based guidelines as a first-line mandatory treatment. However, pragmatic trials and observational studies are needed to determine its real-world effectiveness, impact on outcomes deemed important to patients and to clarify potential mediators of any benefits. METHODS: Individuals with BD were offered the opportunity to participate in 20h of PE and asked to complete pre- and post-intervention ratings of symptoms, knowledge about BD, medication adherence, and illness perception. A priori, two key patient outcomes were identified (social functioning and self-esteem); sample attrition due to dropout or relapse was recorded. RESULTS: Of 156 individuals who completed the pre-PE assessments, 103 completed the program and post-PE assessments. Only 4 of 53 dropouts were associated with BD relapse. Post-intervention, the PE completers demonstrated a statistically significant improvement in social functioning (p = 0.003, Effect Size (ES) = 0.26) and a trend towards improved self-esteem (ES = 0.14). Whilst there were significant changes in medication adherence (p = 0.002, ES = 0.28), knowledge of BD (p < 0.001, ES = 1.20), and illness perception (p < 0.001, ES = -0.37), mediational analysis demonstrated that only change in illness perception was associated to change in functioning (p=0.03) with no contribution from changes in knowledge of BD or medication adherence. CONCLUSIONS: In real-world settings, over 60% individuals completed 10-session course of PE. After controlling for demography and baseline clinical state, change in illness perception, rather than change in knowledge or medication adherence, emerged as a potential mediator of some benefits of PE.

After shrinking dramatically during World War Two the gender wage gap (GWG) narrowed again in the early 1970s due to the Equal Pay Act. The GWG has closed across birth cohorts at all points in the adult life-cycle but remains. Within birth cohort it rises to middle age before falling again. Among those born in 1958, the raw GWG was 16 percentage points among workers aged 23, rising to 35 percentage points at 42. Among those born in 1970 the gaps were 9 and 31 percentage points at age 26 and age 42 respectively. Differences in men's and women's work experience in mid-life account for much but not all of the raw gap in both cohorts. The GWG is a little larger early in the life cycle when accounting for non-random selection into employment but selection plays no role later in life. Policy options for closing the remaining gap are considered.

Introduction Perinatal and later postnatal adversity (e.g., child sexual abuse) are predictors of psychopathology across the lifespan. However, little is known about the impact of the joint effects of perinatal and postnatal adversity on the longitudinal trajectories of mental health problems from adolescence through adulthood. Method We utilized data from a prospective, longitudinal birth cohort of extremely low birth weight (ELBW; 2500 g) control participants. Self-report data on internalizing (depression, anxiety) and externalizing (antisocial) problems were collected at 12–16, 22–26, and 30–35 years of age. Results A birth weight by child sexual abuse (CSA) interaction was observed such that ELBW survivors exposed to CSA had higher levels of internalizing problems from adolescence through adulthood than NBW participants exposed to CSA. Differences remained significant after adjustment for covariates. Likewise, ELBW survivors exposed to CSA had higher levels of internalizing problems from adolescence through adulthood than ELBW participants who were not exposed to CSA. Limitations Findings are limited by sample attrition due to the longitudinal nature of the study spanning over 30 years as well as the retrospective nature of child sexual abuse reporting. Conclusions Exposure to both perinatal and later postnatal adversity leads to persistently higher internalizing problems than exposure to either adversity alone over more than two decades. These findings suggest that individuals exposed to perinatal adversity may be especially vulnerable to, and persistently affected by, childhood adversity, particularly in the form of depression and anxiety.

Bt (Bacillus thuringiensis) cotton has been commercially grown in India for the past 19 years. Genetic Engineering Approval Committee (GEAC) approved the release of Bt cotton for commercial cultivation in 2002 in western and southern parts of the country. In Punjab, Bt cotton was released for cultivation in 2005. Before the release, Bt cotton was adopted by 72% farmers on 22% of the total cotton area. However, a lot of questions have been raised recently on the impacts of Bt cotton. To find out the long-term socio-economic and environmental impacts of Bt cotton cultivation on cleaner production, we revisited cotton growers surveyed in 2003 and 2004. Before-after, with-without, and difference-in-differences (with and without sample attrition) within farm comparisons were analyzed to find the impact of Bt cotton over time. Our results show that sucking insect pests have replaced bollworms as the key pests. There has been a steep decline in insecticide applications to control bollworms, the target pest of Bt cotton, by 97%; however, this has been offset by an increase in the insecticide application by 154% to control sucking pests. Moreover, the increase in pesticide use was driven by the use of fungicides, which were not applied in cotton earlier, and increased use of herbicides. Our results show overall positive impact of Bt cotton on volume of insecticide active ingredients (a.i.) applied, insecticide applications, use of highly hazardous and riskiest insecticides, and resultant environmental impact of the field use of insecticides on cotton. Yields have stabilized after the commercialization of Bt cotton.

In this paper, we return to the question raised in Dahlberg et al. (2012) concerning a causal relationship between ethnic diversity and preferences for redistribution. A re-analysis of their study indicates that results are based on an endogenous instrument and severe sample attrition bias. Correcting for either of these two problems reveals that there is no relationship between ethnic diversity and preferences for redistribution. More generally, we provide results that put into question the conventional description of the Swedish refugee placement policy.

The aim of this methodological evaluation study is to examine the likely effects of sample attrition on the social and psychological outcomes variables often examined in studies of school leavers and which have been included in many studies of employment. We examine the effects of sample attrition based on analysis of a recently completed 10-year longitudinal study of school leavers The results showed that the dropouts differed from the stayers on a number of variables. For example, the dropouts tended to be older and were more likely to use illicit substances, such as marijuana. Males were more likely to drop out than females. On the other hand, the dropouts did not differ from the stayers in terms of the main variables of interest: psychological well-being (including self-esteem, suicidal ideation, life satisfaction, and other related variables) Given that longitudinal studies of adolescents are often used to inform theoretical developments and policies relating to early intervention and longer term outcomes, these findings provide cautious optimism that selective attrition does not pose a major threat to the external or internality validity of employment studies investigating psychological or social outcomes, but it does raise questions about the importance of dealing with missing data in studies of high-risk behaviours Refereed/Peer-reviewed

This article estimates a dynamic reduced-form model of intra-firm promotions using an employer-employee panel of over 300 of the largest corporations in the U.S. in the period from 1981 to 1988. The estimation conditions on unobserved individual heterogeneity and allows for both an endogenous initial condition and sample attrition linked to individual heterogeneity in demonstrating the relative importance of variables that influence promotion. The role of the executive's functional area in promotion is considered along with the existence and source of promotion fast tracks. We find that while the principal determinant of promotions is unobserved individual heterogeneity, functional area has a high explanatory power, resulting in promotion probabilities that differ by functional area for executives at the same reporting level and firm. No evidence is found that an executive's recent speed of advancement in pay grade has a causal impact on in- sample promotions after conditioning on the executive's career speed of advancement. For high-level executives, fast tracks appear to result from heterogeneity in persistent individual characteristics, not from an inherent benefit in recent advancement itself.

This study investigates how social exclusion, defined as a multidimensional concept encompassing financial deprivation, social isolation, and a lack of basic social rights, is associated with cognitive impairment of elderly Chinese. Using three waves of the Chinese Longitudinal Healthy Longevity Survey (N = 10,923), we find that rural elderly women are the most vulnerable to social exclusion (17.28%) and cognitive impairment (18.52%) among the Chinese elderly. After addressing the methodological challenges of endogeneity and sample attrition due to mortality, and holding an array of demographic, lifestyle, and health variables constant, we find that social exclusion is significantly associated with higher risks of cognitive impairment. Rural elderly women suffering from severe or extreme social exclusion are 23 times more likely to be cognitively impaired as compared to their urban male counterparts who are not socially excluded. Lack of participation in social activity is the most common risk factor associated with cognitive impairment across groups, after controlling for relevant covariates.

Background: Frailty constitutes an important risk factor for adverse outcomes among older adults. In longitudinal studies on frailty, selective sample attrition may threaten the validity of results. Objective: To assess the impact of sample attrition on frailty index trajectories and gaps related to socio-economic status (education) therein among older adults in Europe. Methods: A total of 64,143 observations from 21,044 respondents (50+) from the Survey of Health, Ageing and Retirement in Europe across 12 years of follow-up (2004–2015) and subject to substantial sample attrition (59%) were analysed. We compared results of a standard linear mixed model assuming missing at random (MAR) sample attrition with a joint model assuming missing not at random sample attrition. Results: Estimated frailty trajectories of both the mixed and joint models were identical up to an age of 80 years, above which modest underestimation occurred when a standard linear mixed model was used rather than a joint model. The latter effect was larger for men than women. Substantial education-based inequality in frailty continued throughout old age in both the mixed and joint models. Conclusion: Linear mixed models assuming MAR sample attrition provided good estimates of frailty trajectories up until high age. Thus, the validity of existing studies estimating frailty trajectories based on standard linear mixed models seems not threatened by substantial sample attrition.

The Candrive II cohort researchers have followed a convenient sample of older drivers, aged 70 and older, for five to seven years. One of the goals of this study consists in developing a risk stratification tool that would help identify unsafe older drivers. The validity of such tools depends on how representative the study sample is. We have demonstrated that the Candrive II sample at baseline was representative of older Canadian driver through demonstration of equivalence on variables extracted from the Canadian Community Health Survey – Healthy Aging (CCHS-HA). At baseline, 928 older drivers (mean age = 76.21 5) volunteered in the Candrive II study with 583 of them remaining 5 years later (mean age = 79.8,). We make again use of the equivalence testing approach to compare Candrive II sample at year 5 to CCHS-HA drivers of the same age.

We examine how the criteria for choosing estimation samples affect the ability to detect discretionary accruals, using several variants of the Jones (1991) model. Researchers commonly estimate accruals models in cross-section, and define the estimation sample as all firms in the same industry. We examine whether firm size performs at least as well as industry membership as the criterion for selecting estimation samples. For U.S. data, we find estimation samples based on similarity in lagged assets perform at least as well as estimation samples based on industry membership at detecting discretionary accruals, both in simulations with seeded accruals between 2% and 100% of total assets and in tests examining restatement data and AAER data. For non-U.S. data, we find industry-based estimation samples result in significant sample attrition and estimation samples based on lagged assets perform at least as well as estimation samples based on industry membership, both in simulations and in tests examining German restatement data, with substantially less sample attrition.

The main focus of this paper is to explore the potential for improving econometric specification in modeling hedge fund returns. Specifically, we examine the effects of (1) correcting for selectivity bias due to sample attrition; (2) allowing for nonlinearity; and (3) controlling for fund-specific unobserved heterogeneity. Diagnostic tests confirm the importance of these complications. Using data covering 1996–2008, we show that when selectivity, nonlinearity, and fund heterogeneity are taken into account, we obtain more robust estimates of the effect of key variables on hedge fund returns and managerial efficiency, and the explanatory power of the model is significantly improved. Copyright © 2013 John Wiley & Sons, Ltd.

Objectives To systematically review the literature on the "all-on-four" treatment concept regarding its indications, surgical procedures, prosthetic protocols and technical and biological complications after at least three years in function. Study design The three major electronic databases were screened: MEDLINE (via PubMed), EMBASE, and the Cochrane Library of the Cochrane Collaboration (CENTRAL). In addition, electronic screening was made of the 'grey literature' using the System for Information on Grey Literature in Europe - Open Grey, covering the period from January 2005 up to and including April 2016. Results A total of 728 articles were obtained from the initial screening process. Of these articles, 24 fulfilled the inclusion criteria. Methodological quality assessment showed sample size calculation to be reported by only one study, and follow-up did not include a large number of participants - a fact that may introduce bias and lead to misleading interpretations of the study results. Conclusions The all-on-four treatment concept offers a predictable way to treat the atrophic jaw in patients that do not prefer regenerative procedures, which increase morbidity and the treatment fees. The results obtained indicate a survival rate for more than 24 months of 99.8%. However, current evidence is limited due the scarcity of information referred to methodological quality, a lack of adequate follow-up, and sample attrition. Biological complications (e.g., peri-implantitis) are reported in few patients after a mean follow-up of two years. Adequate definition of the success / survival criteria is thus necessary, due the high prevalence of peri-implant diseases. Key words:All-on-four, all-on-4, tilted implants, dental prostheses, immediate loading.

Background: To limit the influence of attrition bias in assessments of intervention effectiveness, several federal evidence reviews have established a standard for acceptable levels of sample attrition in randomized controlled trials. These evidence reviews include the What Works Clearinghouse (WWC), the Home Visiting Evidence of Effectiveness Review, and the Teen Pregnancy Prevention Evidence Review. We believe the WWC attrition standard may constitute the first use of model-based, empirically supported bounds on attrition bias in the context of a federally sponsored systematic evidence review. Meeting the WWC attrition standard (or one of the attrition standards based on the WWC standard) is now an important consideration for researchers conducting studies that could potentially be reviewed by the WWC (or other evidence reviews). Objectives: The purpose of this article is to explain the WWC attrition model, how that model is used to establish attrition bounds, and to assess the sensitivity of attrition bounds to key parameter values. Research Design: Results are based on equations derived in the article and values generated by applying those equations to a range of parameter values. Results: The authors find that the attrition boundaries are more sensitive to the maximum level of bias that an evidence review is willing to tolerate than to other parameters in the attrition model. Conclusions: The authors conclude that the most productive refinements to existing attrition standards may be with respect to the definition of “maximum tolerable bias.”

Data from the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) was used with the objective of evaluating factors associated with antipsychotic dose changes. CATIE was a randomized, comparative effectiveness trial for patients with schizophrenia in which a clinician prescribed a double-blind, flexible dose of one of five antipsychotic medications. The mean number of capsules prescribed monthly was evaluated to identify the period following up-titration with the least sample attrition in which dose changes were likely in response to clinical factors. Demographic, efficacy and tolerability variables were evaluated in two regression models predicting a one capsule dose decrease or increase. The mean dose increased to 2.7 capsules over the first 3 months. The post-titration plateau was identified as between 3 and 6 months. Factors associated with dose increases included the Clinical Global Impression Scale at 6 months and change in the Positive and Negative Syndrome Scale between 3 and 6 months. Decreased dosing was associated only with lower patient rated at 6 months global impression of health. Neither tolerability measures, nor body weight was significantly associated with dose changes. In conclusion, dose changes were weakly associated with measures of current or changing clinical status and not affected by measures of side effects.

Sample attrition, a constant concern in longitudinal studies, occurs when the burden of participation begins to outweigh the rewards (Apodaca et al. 1998). Over the past several years, an increasing number of facilities in the annual National Survey of Substance Abuse Treatment Services (N-SSATS) have begun complaining about the burden of yearly participation. In response, Mathematica imbedded an experiment in the 2008 N-SSATS that tested whether prefilling some fields in their annual questionnaire would encourage continued participation. The appeal of such an option was unclear since prefilled questions must still be read and updated, if incorrect. N-SSATS is sponsored by the Substance Abuse and Mental Health Services Administration (SAMHSA). The project surveys facilities that provide substance abuse treatment services, approximately 17,000 facilities annually.

In this paper we evaluate the quality of survey data collected by the Malawi Diffusion and Ideational Change Project by investigating four potential sources of bias: sample representativeness, interviewer effects, response unreliability, and sample attrition. We discuss the results of our analysis and implications of our findings for the collection of data in similar contexts.

To examine how methodological choices and artifacts of evaluations influence the estimated effect size of media campaigns on health behavior, we meta-analyzed 44 mediated health campaigns. Effect sizes were largest when the posttest evaluation occurred close to the end of the campaign and utilized a one-group partition-by-exposure design. Use of a self-selected sample and greater sample attrition biased the results by inflating effect sizes. Improvements in the control community over time made it more difficult to detect intervention effect sizes. Design factors that did not result in different effect sizes included panel versus cross-sectional, behavioral versus physiological measure, and nonequivalence in the control and intervention communities at baseline. The results may be useful for evaluators and researchers who seek to understand the impact of methodological factors on the interpretation of campaign effects.

Life course study relies for its empirical basis on the longitudinal surveys that supply the means of charting the development of individuals through their lives. Understanding the institutionalisation of life course study is thus inextricably linked with uncovering the reasons for the development of longitudinal inquiry. The chapter starts with the origins of the concept of ‘life course’ then moves to a brief historical overview of first the early (pre-war) longitudinal studies, then the substantial expansion of studies (post war), and finally the post-millennium contemporary studies, attracting major public investment to build longitudinal “research resources.” The reasons for the expansion are sought in the three areas of technological, methodological, and theoretical change. A discussion of the consolidation and then institutionalisation of the field of study follows leading to an examination of the origins of first the Journal and then the Society for Longitudinal and Life Course Studies. Future prospects are finally considered together with such major challenges as sample attrition, data protection, and the development of longitudinal inquiry to meet contemporary circumstances and needs.

Sample attrition, a major concern in any longitudinal study, is even more problematic when adolescents are the study population. The need to minimize the loss of participants in order to maintain the integrity of the cohort is vital in substance abuse prevention evaluations. The Adolescent Substance Abuse Prevention Study (ASAPS), a national school based prevention study being conducted in six sites across the country, followed students attending schools randomly assigned to either a treatment or control condition from the seventh through the eleventh grade. The percent of study dropouts after the ninth grade pretest warranted consideration of an attrition study to acquire data from those students who were no longer in a study school. A pilot study conducted with a sample of attrition cases in one site (51 of 495) was undertaken to assess the feasibility and costs of a full study. This paper describes the experiences gained from this pilot study in tracking study subjects. An exhaustive protocol was developed and contacts with parents were initiated through telephone calls and flyers sent by mail. Online public records and telephone directories were used to acquire additional contact information. Contact was established with 56.8% of the parents and resulted in completed surveys from 19.6% of the sample at a cost of $1,949.30 per survey.

In large-scale educational assessments such as the Third International Mathematics and Sciences Study (TIMSS) or the Program for International Student Assessment (PISA), sizeable numbers of test administrators (TAs) are needed to conduct the assessment sessions in the participating schools. TA training sessions are run and administration manuals are compiled with the aim of ensuring standardized, comparable, assessment situations in all student groups. To date, however, there has been no empirical investigation of the effectiveness of these standardizing efforts. In the present article, we probe for systematic TA effects on mathematics achievement and sample attrition in a student achievement study. Multilevel analyses for cross-classified data using Markov Chain Monte Carlo (MCMC) procedures were performed to separate the variance that can be attributed to differences between schools from the variance associated with TAs. After controlling for school effects, only a very small, nonsignificant proportion of the variance in mathematics scores and response behavior was attributable to the TAs (< 1%). We discuss practical implications of these findings for the deployment of TAs in educational assessments.

This paper asks whether the degree to which our incomes resemble our parents’ incomes has changed over time. After comparing various methods of correcting for differences in the ages at which successive birth cohorts are observed, and correcting for sample attrition, I find no evidence of a linear trend in the intergenerational elasticity of family income for those born into the Panel Study of Income Dynamics between 1952 and 1975 and observed as adults between 1977 and 2000.

The implementation of panel studies in developing countries can prove valuable in driving evidenced-based policies. Case analysis from panel studies in developing countries are therefore useful in offering guidance and insight to deal with the immense challenges of implementing and following-up a panel study in a developing country. The aim of this paper is to describe the follow-up process within Birth to Twenty (Bt20), a birth cohort study in Johannesburg, South Africa, over the past 16 years; and to outline retention strategies and insight within the context of an urban, developing country setting. The main issues considered are: attrition (periodic and absolute), movement of individuals, name changes, cohort management and lessons learned. Copyright © 2007 John Wiley & Sons, Ltd.

The data of the Russian Longitudinal Monitoring Survey (RLMS) – Higher School of Economics represents one of the few nationally representative sources of household and individual data for Russia. These data have been collected since 1992 and in recent years, thanks to more secure financial and logistical support, have become a resource increasingly drawn upon by scholars and students for national and cross‐national studies. In this paper, we examine the extent of non‐random attrition in the RLMS and discuss the circumstances under which this might give rise to biases in econometric analysis. We illustrate this with an example drawn from the health sphere.