This is a test by dfhdhdhdhjr


									Review of Statistics for
       Pulmonary Lab

            C. S. Tritt, Ph.D.
            April 11, 2006
Why Bother with Statistics?
 There   is always some uncertainty
  in measured values.
 It is often impractical to sample
  an entire population, so some
  subset must be selected and
  measured and the result used to
  draw a conclusion about the
Why Bother with Statistics?
 Statistics quantifies the inherent
  uncertainties of experimental
 Statistics allows uncertainty to be
  quantified. This can prevent
  disputes in the interpretation of

 Nearly all experimental
  investigations should start with
  one or more hypothesis.
 Statistical methods are often used
  to test these hypothesis.
 This lab provides a specific
  example of this type of use of
Test Statistics
 In hypothesis testing, a test statistic
  is typically calculated from the
  observed data.
 This value is then compared to an
  expected value based on a
  mathematical analysis of the general
 If the test statistic is greater than
  the expected value, reject the
  hypothesis than your sample
  matches the expected situation.
Comparison of Means
 In this lab, you were asked to
  compare your results with accepted
  values of certain pulmonary
 These volumes and there values
     TV = 500 ml
     IRV = 1900 ml
     ERV = 700 ml

   In general, t-test’s are used for the
    comparison of means.
TV, IRV & ERV Instructions
 Excel has some a built in data analysis
  tools. These are listed under the Tools
  | Data analysis menu choice.
 However, it do not appear to be able
  to make comparisons between
  samples and known values so some
  manual calculations were required.
 Alternatively, you may be able to use
  Minitab or some other statistical
  analyses software to do your analysis.
VC Instructions
 In this experiment you will be
  comparing observed and predicted
  VC values.
 I suggest you use unpaired t-tests
  (although paired tests might be
  better) not assuming equal
  variances to do this.
 In this case, since you have a
  different predicted value for each
  subject, the variance of the
  predicted values can be calculated.
    My Results (part 1)
 See NursesPulmVolResultWork.xls.
 Note use of sheet labels.
 Note layout of table with calculated
  (function) values.
 Note entry of expected values for
  TV, IRV and ERV.
 Note use of prediction equation.

My Results (part 2)
 Note location and labeling of
  statistical test results.
 Note identity plot for VC.
 Note formatting of numeric values
  (significant digits).
 Note formatting of gridlines, axis
  labels and initials.
 Did not do RV calculations.
 Did not use paired t-tests.


To top