China
Africa
Explore chapters and articles related to this topic
The Study Population:
Published in Lynne M. Bianchi, Research during Medical Residency, 2022
Lynne M. Bianchi, Luke J. Rosielle
Volunteer and non-responder biases are the two primary forms of participant bias that nearly every study must consider. Volunteer bias reflects the fact that individuals who choose to participate in a study are different in some ways from those who do not volunteer. Similarly, those who do not participate (non-responders) are different in some ways from those who do. Non-response bias includes those who decline to participate, those who are difficult to reach, and those who fail to follow-up once enrolled in a study.
Strategies to Handle Missing Data in Meta-Analysis
Published in Ding-Geng (Din) Chen, Karl E. Peace, Applied Meta-Analysis with R and Stata, 2021
Informative missingness (IM) approach is a unique way to handle missing observations in meta-analysis especially in aggregate (summary) data from published studies. When aggregate data in meta-analysis are MNAR but they are analyzed under the assumption of MAR or MCAR then nonresponse bias can occur. White et al. (2008) proposed the method of IM for binary outcomes with aggregate data by quantifying the ratio of the odds of the outcome among subjects with unobserved outcome to the odds of the outcome among observed subjects. They referred this quantity as informative missingness odds ratios (IMORs). Mavridis et al. (2015) extended the IM approach to meta-analyses with continuous outcomes by defining informative missingness parameters (IMPs) that relate the mean of the outcome between the missing and the observed participants. IMPs are like sensitivity parameters which measure the departure from MAR assumption (Kenward et al., 2001). Many statistical softwares such as STATA can perform sensitivity analysis for aggregate data (Chaimani et al., 2018). In our next section, we will show how to use IMORs to analyze binary outcome data for meta-analyses.
Contact Patterns for Contagious Diseases
Published in Leonhard Held, Niel Hens, Philip O’Neill, Jacco Wallinga, Handbook of Infectious Disease Data Analysis, 2019
Jacco Wallinga, Jan van de Kassteele, Niel Hens
An alternative approach to deal with the identifiability problem is to use more data, of a different kind: we can use information from social behavior surveys to estimate the contact rates. This use should be done with great care, as information from behavior surveys can be subject to difficulties associated with nonresponse bias, insufficient number of participants, and untruthful answers by respondents. The use of surveys on sexual behavior to inform contact rates in models for sexually transmitted infections, such as infections with gonorrhea and HIV, dates back to the 1990s [5, 6]. The development of social contact behavior surveys to inform contact rates for models of close contact infections or airborne droplet infections, such as measles and influenza, is more recent [1, 7, 8].
An intersectional investigation of Asian American men’s muscularity-oriented disordered eating: Associations with gendered racism and masculine norms
Published in Eating Disorders, 2022
Thomas P. Le, Benjamin T. Bradshaw, M Pease, Linda Kuo
Furthermore, participation bias and non-response bias limit the generalizability of our results. Only Asian American men who had access to some form of working internet could have completed our study’s survey, and there may be additional factors that predisposed certain men to choose to participate in our study as well as factors that may have disincentivized participation. For example, we advertised through universities’ Asian cultural organizations and various social media sites, which means that our study’s results may not generalize to Asian American men who have not participated in cultural organizations or who do not use social media. Additionally, based on the power analysis, the current study lacked a sample size sufficient to detect small effects, leaving the possibility that null results are Type II errors (i.e., failing to detect a small, but true, effect). Continued research in this area with larger samples is necessary to elucidate all factors relating to Asian American men’s MODE. Lastly, it is also important to note that the modified EDE-Q we used in this study has not been validated as a measure of MODE. Although this modified measure has shown high internal reliability in previous studies (Griffiths et al., 2015; Murray et al., 2012) and in the current study, caution should be used when interpreting the results based on this measure given its lack of validation.
Effectiveness of an Aggression Management Training Program in Japan: A Quasi-Experimental Study
Published in Issues in Mental Health Nursing, 2022
Aimi Kinoshita,, Seiji Shimosato,
Concerning nonrespondent bias, the number of questions was examined and reduced to limit the burden on respondents. However, given that this was an arbitrary survey, nonrespondents could not be avoided, possibly leading to nonresponse bias. We also asked about the motivation for attending the training, which may have introduced self-selection bias. The numbers for active and passive motivation were almost the same. No significant difference was evident in the outcome. However, this survey used a pretest–posttest design, and training was conducted across various parts of Japan. Therefore, establishing a control group was not possible, which constituted another limitation of this study. Given that no control group was included, we could not determine the magnitude of the effect that may lead to the Hawthorne effect. Although the training was limited because of COVID-19, including a control group in the future would be helpful.
Relationship Between CT Head Findings and Long-term Recovery in Children with Complicated Mild Traumatic Brain Injury
Published in Brain Injury, 2022
Colby Hansen, Laura C. Waller, Dalton Brady, Masaru Teramoto
As for limitations, it is important to note this was not a longitudinal study. Consequently, we cannot comment on the evolution of important prospective indicators of recovery, such as repeated symptom scores and quality of life measures. There are several forms of potential bias due to some of the data for this study being collected by retrospective survey. For example, nonresponse bias may be present due to a relatively low response rate (23.2% in this study), although the demographics of the survey responders and nonresponders were not significantly different. An inherent weakness of retrospective studies is recall bias. This study in particular employed surveys relatively far removed from the incident injury (average 3.9 ± 1.1 years). Caregivers could potentially have biased recall of their child’s recovery from an injury, particularly if they experienced other subsequent unrelated health challenges or injuries between the incident injury and time of survey. Additionally, because of the nature of cross-sectional studies, we cannot make causal inferences between potential injury-related predictors and recovery outcomes after C-mTBI. Certainly there are many non-injury factors that could affect recovery in mTBI, such as racial, ethnic, or socioeconomic factors, for which we did not control (11).