What we addressed in this unit

  1. Exploring data – looking at things to see what of interest might strike your eye
    1. Bar charts, histograms, pie charts … each shows something different about the data.
    2. Differentiating between plots applicable for one variable vs. plots app
  2. "Weighting" scores to come up with a ranking system
  3. Comparing weighting schemes to see how closely related they are
  4. Association between variables
    1. "Seeing" it, and comparisons thereof
      1. Presentations of data (e.g., dotplots y by x, side-by-side box plots, scatter plots)
      2. Issues of consistency of scale (percents of largest score, ranks, z-scores)
    2. Measuring association
      1. Sums of differences between ranks
      2. Sums of absolute differences between ranks
      3. Sums of products of ranks
      4. Sums of products of scores
      5. Sums of products of z-scores
    3. Relative association
      1. Comparing measures of association between data sets having different numbers of cases
      2. Divide by maximum possible sum for that number of cases – no matter what the sum is summing.
    4. Project – predicting reputation scores for colleges that haven't been given one.
      1. Find formula of combinations that yield highest possible correlation for colleges having reputation scores using data common to all colleges
      2. Apply that formula to all colleges