How to choose an approach to handling missing categorical data: (un)expected findings from a simulated statistical experiment

Zhuchkova, Svetlana; Ротмистров, Алексей Николаевич

doi:10.1007/s11135-021-01114-w

Cited by 19 publications

(13 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In total, 10,090 children were recruited during the baseline survey (termed Wave 1 by the LSAC) in 2004, and in the following waves, data were gathered from the same participants as they aged from 2004 to 2020. The details of the LSAC study design and data collection procedures are described elsewhere [ 40 , 41 ].…”

Section: Methodsmentioning

confidence: 99%

“…In this study, we used two recent LSAC waves—Wave 9C1 (i.e., data collection conducted during the COVID-19 pandemic between October and December 2020, referred to here as the ‘COVID-19 pandemic’) and Wave 8 (i.e., data collected in 2018 before the COVID-19 pandemic, referred to here as ‘before the COVID-19 pandemic’ or ‘pre-COVID-19’) to include a sample of 1110 respondents only from the older K-cohort (who were aged 20–21 years during the COVID-19 wave, and were aged 18–19 years in the pre-COVID-19 wave) of the LSAC database [ 41 ]. We used K-cohort as only this cohort responded to the same questions regarding healthcare avoidance in both the 2018 and 2020 data waves.…”

Section: Methodsmentioning

confidence: 99%

“…In this study, we listed the variables in Table 1 that were deemed to be potential determinants of healthcare seeking or avoidance in accordance with previous studies conducted among young people [ 48 , 49 , 50 ]. In addition, we included variables related to the COVID-19 pandemic including the coronavirus restriction period (CRP) (also known as ‘lockdown’) between March-May 2020, and these were only available from the COVID-19 wave (i.e., LSAC Wave 9C1 data) [ 41 ].…”

Section: Methodsmentioning

confidence: 99%

See 2 more Smart Citations

Healthcare Avoidance before and during the COVID-19 Pandemic among Australian Youth: A Longitudinal Study

et al. 2022

View full text Add to dashboard Cite

Background: Access to healthcare for young people is essential to ensure they can build a foundation for a healthy life. However, during the COVID-19 pandemic, many people avoided seeking healthcare, adversely affecting population health. We investigated the factors associated with the avoidance of healthcare for Australian young people when they reported that they needed healthcare. We were able to compare healthcare avoidance during the COVID-19 pandemic with healthcare avoidance prior to COVID-19. Methods: We used two recent data collection waves from the Longitudinal Study of Australian Children (LSAC)—Wave 9C1 during the COVID-19 pandemic in 2020, and Wave 8 data which were collected in 2018. The primary outcome of this study revealed the avoidance of healthcare among those who perceived the need for care. Bivariate analyses and multiple logistic regression models were employed to identify the factors associated with the avoidance of healthcare during the COVID-19 and pre-COVID-19 periods. Results: In the sample of 1110 young people, 39.6% avoided healthcare during the first year of the COVID-19 pandemic even though they perceived that they had a health problem that required healthcare. This healthcare avoidance was similar to the healthcare avoidance in the pre-COVID-19 pandemic period (41.4%). The factors most strongly associated with healthcare avoidance during the COVID-19 pandemic were female gender, an ongoing medical condition, and moderately high psychological distress. In comparison, prior to the pandemic, the factor associated with healthcare avoidance was only psychological distress. The most common reason for not seeking healthcare was thinking that the problem would spontaneously resolve itself (55.9% during COVID-19 vs. 35.7% pre-COVID-19 pandemic). Conclusions: A large proportion of youths avoided healthcare when they felt they needed to seek care, both during and before the COVID-19 pandemic.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Healthcare Avoidance before and during the COVID-19 Pandemic among Australian Youth: A Longitudinal Study

et al. 2022

View full text Add to dashboard Cite

show abstract

“…A complete case analysis (CCA) was performed since data were missing completely at random and were representative of the entire study sample. Moreover, it has been reported that typically CCA produces unbiased results in regression models as it works with actual data rather than simulated data [52].…”

Section: Participantsmentioning

confidence: 99%

Bullied Because of Their Teeth: Evidence from a Longitudinal Study on the Impact of Oral Health on Bullying Victimization Among Australian Indigenous Children

Islam¹,

Chadwick²,

Esgin³

et al. 2022

Preprint

View full text Add to dashboard Cite

Making life better for Indigenous peoples is a global priority. Although bullying and oral health have always been a topic of concern, there is limited information regarding the impact of this problem in the general population, with no evidence in this regard among the Australian Indigenous population. Thus, we aimed to quantify the relationship between bullying victimization and oral health problems by remoteness among 766 Australian Indigenous children aged between 10–15-year-olds using data from the LSIC study. Bivariate and multilevel mixed-effect logistic regression analyses were employed. Findings indicated children self-reported bullying more than parents reported their children were being bullied (44% vs 33.6%), with a higher percentage from rural/remote areas than urban areas. Parents reported oral health problem increases the probability (OR 2.20, p<0.05) of being bullied in Indigenous children living in urban areas. Racial discrimination, lower level of parental education and poor child oral hygiene increase the risk of bullying victimization. Parental happiness with life and a safe community was associated with a lower risk of bullying. Dental problems are linked with Australian Indigenous children experiencing bullying victimization. Cultural resilience and eliminating discrimination maybe two modifiable paths to ameliorating health issues associated with bullying in the Australian Indigenous community.

show abstract

“…В случае превышения порога в 5 % для перераспре деления неопределившихся респондентов использовался метод множественной импутации, обоснованный в работе [Фабрикант, 2015] и реализуемый в виде методики, подробно рассмотренной в [Воробьев и др., 2020]. В состав методики входят следующие основные процедуры: -выбор независимых переменных, коррелируемых с зависимой переменной («Партия»), с целью приведения неслучайных пропусков (NMAR) к случайным (MAR) с использованием подхода [Enders, 2010] и экспериментально проверенного в работах [Зангиева, 2011;Zhuchkova, Rotmistrov, 2022]; -расчет импутированных значений вместо пропусков с помощью процеду ры «множественная импутация» (метод импутации «автоматический» обоснован в [Ворбьев, Воронецкий, 2019]), реализованной в программе IBM SPSS Statistics версии 23, где в качестве предикторов использовались отобранные в предыдущей процедуре независимые переменные.…”

Section: особенности используемых в работе социологических данныхunclassified

Методика Прогнозирования Результатов Выборов С Использованием Комбинирования Методов Аналитического И Трендового Моделирования

Воробьев¹,

Ноздрачева²

2022

MONITORING OF PUBLIC OPINION

View full text Add to dashboard Cite

Работа посвящена исследованию возможностей прогнозирования рейтинга политических партий на коротких временных рядах с помощью данных социологических опросов ВЦИОМ по выборной тематике. Рассмотрены три основные группы методов: трендовое моделирование, экспертные оценки, аналитическое моделирование. В результате теоретического анализа и проведенных экспериментов выявлена низкая точность прогнозов, полученных только с помощью методов, относящихся к группе трендового моделирования. Поэтому предложена методика, позволяющая повысить точность прогнозирования за счет корректировки полученных в результате логарифмической аппроксимации прогнозов. С помощью иерархической кластеризации находится вектор значений, состоящий из коэффициентов уравнений аппроксимации и описывающий наиболее «похожую» электоральную ситуацию в прошлом. Затем «остаток», вычисленный как разница суммы прогнозных значений и 100%, пропорционально перераспределяется в соответствии с участниками избирательной кампании. Для прогнозирования рейтинга партий, где значения временного ряда не превышают 5%, предпочтительнее применять метод среднего темпа роста.

show abstract

How to choose an approach to handling missing categorical data: (un)expected findings from a simulated statistical experiment

Cited by 19 publications

References 37 publications

Healthcare Avoidance before and during the COVID-19 Pandemic among Australian Youth: A Longitudinal Study

Healthcare Avoidance before and during the COVID-19 Pandemic among Australian Youth: A Longitudinal Study

Bullied Because of Their Teeth: Evidence from a Longitudinal Study on the Impact of Oral Health on Bullying Victimization Among Australian Indigenous Children

Методика Прогнозирования Результатов Выборов С Использованием Комбинирования Методов Аналитического И Трендового Моделирования

Contact Info

Product

Resources

About