Basic Statistics

• What type of comparison can statistics help with?

  Comparing responses of different medicines
• What prediction can statistics assist with?

  Product success with the general public
• What type of data is nominal scale?

  Categorical data without natural ordering
• What is an example of ordinal scale data?

  Patient opinion on a visual analogue scale
• What characterizes nominal scale data?

  No natural ordering, arbitrary sequence
• What is dichotomous data?

  Data with only two categories
• What is a variable in statistics?

  A value that can vary
• What are the two major types of variables?

  Categorical and metric variables
• What is the algorithm for determining variable type?

  1. Has the variable got units?
  2. Can the data be put in meaningful order?
  3. Do the data come from measuring or counting?
• What distinguishes samples from populations?

  Samples test part of a population
• What are the measures of central tendency?

  • Mean: average value
  • Median: middle ranking value
  • Mode: most frequently occurring value
• What is the standard deviation used for?

  Indicator of dispersion in data
• What does deviation about the mean indicate?

  How far each data point is from the mean
• What are the characteristics of normal data?

  • Unimodal
  • Symmetrical
  • No sudden cut-offs
• What does a positively skewed distribution indicate?

  Mean is greater than the median
• What is the significance of the standard error of the mean (SEM)?

  • Estimates sampling error
  • Based on sample size and SD
  • Reflects variability among sample estimates
• What does a 95% confidence interval indicate?

  Range where true population mean lies
• What happens if the confidence interval includes zero?

  Null hypothesis is credible
• What is the null hypothesis?

  No real effect, only random sampling error
• What is the alternative hypothesis?

  There is a real effect causing the difference
• What are the requirements for applying a t-test?

  • Samples must be normally distributed
  • Samples must have equal SDs
• What does the P value indicate?

  Likelihood of observing an effect by chance
• How do P values and confidence intervals relate to significance?

  • Both assess evidence of an effect
  • P value answers "is there an effect?"
  • CI answers "how great is the effect?"