Scholarlink

<p>Non-response bias threatens the validity of mental health survey research, yet its influence on estimating psychopathology remains underexplored. Drawing on epidemiological methods within a psychological framework, this study explores non-response patterns within <i>The Aarhus Bereavement Study</i>, a cohort of 2,006 bereaved spouses identified through Danish registers. We distinguished two main types of non-response: (1) complete non-response (uncontactable individuals, survey refusers, never-responders) and (2) partial non-response (follow-up dropouts, missing key items)–and examined for socio-demographic and physical health predictors. Using adjusted regression models, we found that complete non-responders were more likely to be older (&gt; 75&#xa0;years old). In contrast, partial non-responders were younger (&lt; 44&#xa0;years old), more likely to have only one child, have poorer physical health, and stable tobacco use, while less likely to have a higher education and less likely to consume alcohol. Importantly, baseline poor physical health predicted both later partial non-response (OR = 1.91; 95% CI [1.41, 2.59]) and follow-up psychopathology among complete cases (OR = 5.02; 95% CI [2.70, 9.31]), suggesting a latent confounding factor. These results demonstrate that partial non-response is not random but systematically associated with key risk factors for poor mental health. This non-randomness can introduce substantial estimation bias, potentially obscuring the true prevalence of grief-related distress. We argue for the inclusion of methodological indicators in psychological study research to assess and adjust for non-response bias, bridging epidemiological rigor with psychological theory to improve the empirical validity of mental health survey outcomes.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Recruitment and retainment of bereaved spouses: non-response to a Danish registry-sampled longitudinal survey on mental health after loss

  • 期刊论文

  • 发表日期：

    2025年12月20日

查看全文
  • Madeline Marie Marello,
  • Christy Denckla,
  • Maja O’Connor

摘要

Non-response bias threatens the validity of mental health survey research, yet its influence on estimating psychopathology remains underexplored. Drawing on epidemiological methods within a psychological framework, this study explores non-response patterns within The Aarhus Bereavement Study, a cohort of 2,006 bereaved spouses identified through Danish registers. We distinguished two main types of non-response: (1) complete non-response (uncontactable individuals, survey refusers, never-responders) and (2) partial non-response (follow-up dropouts, missing key items)–and examined for socio-demographic and physical health predictors. Using adjusted regression models, we found that complete non-responders were more likely to be older (> 75 years old). In contrast, partial non-responders were younger (< 44 years old), more likely to have only one child, have poorer physical health, and stable tobacco use, while less likely to have a higher education and less likely to consume alcohol. Importantly, baseline poor physical health predicted both later partial non-response (OR = 1.91; 95% CI [1.41, 2.59]) and follow-up psychopathology among complete cases (OR = 5.02; 95% CI [2.70, 9.31]), suggesting a latent confounding factor. These results demonstrate that partial non-response is not random but systematically associated with key risk factors for poor mental health. This non-randomness can introduce substantial estimation bias, potentially obscuring the true prevalence of grief-related distress. We argue for the inclusion of methodological indicators in psychological study research to assess and adjust for non-response bias, bridging epidemiological rigor with psychological theory to improve the empirical validity of mental health survey outcomes.