Explanation:

Detection bias, also known as ascertainment bias or observer bias, occurs when the assessment of the outcome is systematically different between the groups being compared, often due to the assessor's knowledge of the intervention. If the raters who assess the outcome are not blinded to the treatment allocation, their knowledge could influence how they look for, interpret, or record the outcomes, leading to biased results.

Explanation:

When a study has multiple outcomes (a primary outcome and secondary outcomes), and only statistically significant secondary outcomes are highlighted while a non-significant primary outcome is downplayed or less emphasized, this can be an indication of reporting bias, sometimes referred to as "outcome reporting bias." It involves selectively reporting results that are statistically significant or favorable, which can distort the overall findings and conclusions drawn from the study.

Explanation:

Internal validity concerns whether the observed effect in a study is truly due to the exposure or intervention being investigated, rather than to other factors. Confounding occurs when a third variable is associated with both the exposure and the outcome and distorts the true relationship between them. Ensuring internal validity requires minimizing bias and controlling for confounding factors so that the observed association is likely a true one within the study population.

Explanation:

Intention-to-treat analysis is a method used in RCTs where participants are analyzed in the group they were originally randomized to, regardless of whether they completed the intervention or switched groups. This approach maintains the comparability of the treatment groups established by randomization, which is crucial for minimizing selection bias and other forms of bias that can occur after randomization, thereby preserving the integrity of the study design.

Explanation:

Attrition bias occurs in studies that follow participants over time (longitudinal studies) if participants drop out differentially based on exposure or outcome status. Cross-sectional studies are not longitudinal; they collect data at a single point in time. Since there is no follow-up period, participants do not drop out during the study, making cross-sectional designs inherently resistant to attrition bias during the data collection phase.

Explanation:

Reporting bias occurs when the dissemination of research findings is influenced by the nature or direction of the results. Specifically, it can involve selectively reporting outcomes or adverse events. Underreporting unfavorable results, like adverse events, is a classic example of reporting bias, where the information presented in the final report is skewed.