Power of largest root on canonical correlation

The analysis of covariance provides a common approach to adjusting for a baseline covariate in medical research. With Gaussian errors, adding random covariates does not change either the theory or the computations of general linear model data analysis. However, adding random covariates does change the theory and computation of power analysis. Many data analysts fail to fully account for this complication in planning a study. We present our results in five parts. (i) A review of published results helps document the importance of the problem and the limitations of available methods. (ii) A taxonomy for general linear multivariate models and hypotheses allows identifying a particular problem. (iii) We describe how random covariates introduce the need to consider quantiles and conditional values of power. (iv) We provide new exact and approximate methods for power analysis of a range of multivariate models with a Gaussian baseline covariate, for both small and large samples. The new results apply to the Hotelling-Lawley test and the four tests in the "univariate" approach to repeated measures (unadjusted, Huynh-Feldt, Geisser-Greenhouse, Box). The techniques allow rapid calculation and an interactive, graphical approach to sample size choice. (v) Calculating power for a clinical trial of a treatment for increasing bone density illustrates the new methods. We particularly recommend using quantile power with a new Satterthwaite-style approximation.

show abstract

“…Most results involve asymptotic properties and use zonal polynomial forms [6][7][8][9][10][11][12][13][14][15][16].…”

Section: Literature Reviewmentioning

confidence: 99%

Adjusting power for a baseline covariate in linear models

Glueck

Muller

2003

Statistics in Medicine

View full text Add to dashboard Cite

show abstract

“…Specifically, participant age was negatively correlated with signal memory and positively correlated with signal kurtosis. The power of the final test was estimated as lying between 70 and 80% according to the tables from Sugiyama and Ushizawa [41].…”

Section: Resultsmentioning

confidence: 99%

“…The power of the canonical correlation test for zero correlation was estimated based on the formulation of Sugiyama and Ushizawa [41], who reported power estimates for canonical problems comparable to those considered here in terms of the number of variables, the strength of the correlations, and the sample size.…”

Section: Discussionmentioning

confidence: 99%

Anthropometric and Demographic Correlates of Dual-Axis Swallowing Accelerometry Signal Characteristics: A Canonical Correlation Analysis

et al. 2009

View full text Add to dashboard Cite

Swallowing accelerometry has been proposed as a potential minimally invasive tool for collecting assessment information about swallowing. The first step toward using sounds and signals for dysphagia detection involves characterizing the healthy swallow. The purpose of this article is to explore systematic variations in swallowing accelerometry signals that can be attributed to demographic factors (such as participant gender and age) and anthropometric factors (such as weight and height). Data from 50 healthy participants (25 women and 25 men), ranging in age from 18 to 80 years and with approximately equal distribution across four age groups (18-35, 36-50, 51-65, 66 and older) were analyzed. Anthropometric and demographic variables of interest included participant age, gender, weight, height, body fat percent, neck circumference, and mandibular length. Dual-axis (superior-inferior and anterior-posterior) swallowing accelerometry signals were obtained for five saliva and five water swallows per participant. Several swallowing signal characteristics were derived for each swallowing task, including variance, amplitude distribution skewness, amplitude distribution kurtosis, signal memory, total signal energy, peak energy scale, and peak amplitude. Canonical correlation analysis was performed between the anthropometric/demographic variables and swallowing signal characteristics. No significant linear relationships were identified for saliva swallows or for superior-inferior axis accelerometry signals on water swallows. In the anterior-posterior axis, signal amplitude distribution kurtosis and signal memory were significantly correlated with age (r = 0.52, P = 0.047). These findings suggest that swallowing accelerometry signals may have task-specific associations with demographic (but not anthropometric) factors. Given the limited sample size, our results should be interpreted with caution and replication studies with larger sample sizes are warranted.

show abstract

Power of largest root on canonical correlation

Cited by 2 publications

References 3 publications

Adjusting power for a baseline covariate in linear models

Adjusting power for a baseline covariate in linear models

Anthropometric and Demographic Correlates of Dual-Axis Swallowing Accelerometry Signal Characteristics: A Canonical Correlation Analysis

Contact Info

Product

Resources

About