LS 40 (Life Science) Midterm Exam 1 Study Guide updated in Jan 2021(A+ Guide)

QUESTIONS TO ASK AFTER PLOTTING DATA TO FIND WHAT SIMULATION TO USE What can I notice from the plot (see “What to look for in a data plot” section)? What measure of central tendency would be bes t for this set of data (see “Measures of Central Tendency” section) Are there 1 group or 2 groups? 1 group: 2 groups: Are the two groups independent or paired? Are we calculating P-value or CI? Week 1 ● Step 1: ALWAYS VISUALIZE DATA ● Types of Plots ○ Dot Plots ■ Dots can overlap when ~ 50 data points ○ Beeswarm Plots ■ Stacks overlapping points ■ Allows us to see every single data point ■ Too much data=overwhelming ○ Histograms ■ Sensitive to bin size ● Too large or too small ○ Cannot see distribution ■ y-axis=count ■ Kernel Density Estimates ● Smoothes out the histogram ○ Violin Plot ■ Problem: cannot see the raw data ■ Can see distribution and bimodality ○ Box Plots ■ Show distribution/spread of the data and identifies outliers ■ Shows IQR ● 25th percentile, median, 75th percentile (second ¼ - third ¼) ● Lowest ¼, second ¼, median, third ¼, Highest ¼ ● What to look for in a data plot ○ Central Tendencies ■ Bunching up near the middle? ○ Tails ■ Thick or thin? ○ Distribution ■ Symmetric or skewed? ■ Meanmedian= right skew ■ Meanmedian=left skew ○ Outliers ■ Any extreme values? ○ Subgroups ■ Are there multiple groups? Bimodality ○ Variability ■ How variable are the data? How spread out is it around the central tendency? ● Measures of Central Tendency ○ Mean ■ Strongly affected by extreme values ■ Only use if distribution is bell shaped (gausian) ● Symmetric ● Thin Tailed ● Unimodal ○ Median ■ Largely unaffected by extreme values ■ Median does not equal mean ■ Gives actual information and is more accurate with skewed data