Posted: September 13th, 2017

Biostatistics Hw 8

Paper, Order, or Assignment Requirements

Ill upload the sass files

HW 8 operating characteristics, kappa

1. One theory of chronic myofascial face pain holds that patients brux (grind their teeth) during sleep. However, most evidence supporting this theory comes from patient reports rather than objective assessment. How well do self reports describe actual bruxism? A recent study was done in a sleep lab (124 chronic face pain cases and 46 no pain controls) in order to assess the relationship between a subjects’ reports of bruxism and a gold standard assessment of noctural bruxism (via analysis of the EMG record, audio and visual monitoring during sleep). In your summary of the analyses of the ‘sleep bruxism data’, include the following:
a. Do rates of reported bruxism differ between cases and controls (p< .05)?
b. Do rates of grinding during sleep differ between cases and controls (p< .05)?
c. Do rates of sleep bruxism diagnosis differ between cases and controls (p< .05)?
d. Looking only in the cases, describe the operating characteristics of self-report for the occurrence of sleep sounds and then for diagnosis of sleep bruxism. In your description of the analysis, describe the Sn, Sp, PPV and NPV of the test.
d. What does kappa indicate about the concordance between self report and each of the objective measures?
e. Comment on the utility of self-reported bruxism to the actual diagnosis.

2. A researcher is interested in the reliability of raters who determine the failure mode (from SEM images) that occurs when shearing the bond between a tooth and orthodontic bracket. Two materials scientists are asked to judge the failure mode of 25 samples as showing failure in the cement layer (0), the enamel layer (1), or mixed (2). Use ‘failure mode reliability data’ file.
a. Is the determination of failure mode reliable?
Double the sample size. Is there any effect on reliability?
b. Can this result be generalized to all possible raters?
c. Sometime the unreliability comes from uncertainties in type of rating. What happens to reliability when you collapse the rating scale to cement layer vs. other (enamel or mixed) (ie, recode the original variable to 2 categories)?
D. Summarize the findings in a paragraph.