research thesis proposal written and oral.

You are highly encouraged to use schematics and tables to appropriately consolidate your data used for your sample size calculations, your planned contrasts etc. 1. State simply the proposed appropriate model you will be using for your experiment. You will also want to consider adding in the appropriate assumptions (normally distributed, date will be transformed if necessary, independent observations, etc.) A questionnaire will be developed to predict energy deficiency. The score will be tested for correlations with triiodothyronine concentration, bone mineral density, and resting metabolic rate ratio. A cut-off score for energy deficiency will be established and sensitivity, specificity, positive predictive value, and negative predictive value for low T3 (<80 ng/dL), low bone mineral density (z-score<-1), and low RMRratio (<0.9) will be tested. The variables are expected to be normally distributed, and they will be log-transformed in case they are not. If transformation does not result in a normal distribution, non-parametric tests will be used. 2. Indicate the values you have possibly found from the literature in terms of what potential observed differences you would anticipate. The Low Energy Availability questionnaire is reported to have 78% sensitivity and 90% specificity for low energy availability, low bone mineral density, and/or menstrual dysfunction, with a sample of 45 women. Considering that our study will include a larger sample and three metabolic variables, it is expected that a higher specificity will be found. 3. State your anticipated/desired effect size. A correlation coefficient between the questionnaire score and T3, bone mineral density, and RMRratio is expected to be at least 0.7. 4. Describe what you would consider/determine a meaningful difference you would consider significant in your field of study. A meaningful difference would be a tool with 90% sensitivity, using variables at a 95% CI. 5. Describe the variance you would anticipate in your measurements, and the expected variance in the difference if you are doing an intervention (pre/post measures). A 87 variance for T3, 0.01 for RMRratio, and 9 for questionnaire. 6. For studies utilizing regression or multiple regression describe what you are examining in terms of predictors in your model and the anticipation of the type of model (linear, nonlinear). Questionnaire score is a discrete variable that is expected to predict T3, bone mineral density, and RMRratio. A log distribution is expected. 7. Justify your final sample size to ensure that you will get meaningful data with this design and your anticipated analytical approach. For a questionnaire, 10 subjects per question: n=100.

For validation with sensitivity and specificity using T3: estimated sample size with 95% C.I, SD of 18, and margin of error for confidence interval of 3: 95.