paperKB
coga / coga-kb
Help
Sign in

Chunk #25 — 3. Results — 3.4. Correlations across Significant Predictors

Source
Predicting Alcohol-Related Memory Problems in Older Adults: A Machine Learning Study with Multi-Domain Features.
Embedded
yes

Text

An exploratory (descriptive) analysis of correlations among the top significant variables is shown in Figure 3. As shown in the correction matrix, there were significant positive correlations among the functional connectivity variables within and between different frequency bands. Overall, most of the low-frequency connections in the delta and theta frequencies were highly correlated with one another. Specifically, those connections that shared a common node showed much higher correlations with each other than with other connections, regardless of their frequency bands. Beta band connections had significant positive correlations between themselves as well as with low-frequency connections, especially theta band connections. However, alpha and gamma band connections showed significant correlations only within the frequency and not across the frequencies. Highly significant positive correlations were observed among the alcohol-related health consequences. Among the personality factors, there was a significant positive correlation between neuroticism and harm avoidance. However, no significant correlations were observed across the domains (e.g., functional connectivity vs. personality, or functional connectivity vs. alcohol-related features).