1 of 9

The Use of Statistical Data

in Research

Frank LaBanca, EdD

Connecticut Science Fair Association

2 of 9

Thoughts on the “mean” as a method for decision making

3 of 9

Data Set

inquirylearningcenter.org

4 of 9

Box-and-whisker Plot

Key Data Points:

​

Maximum

75% Quartile (R3)

50% Quartile (Median)

25% Quartile (R1)

Minimum

5 of 9

Box-and-whisker Plot

Great Meadows Marsh

Barn Island Marsh

Sherwood Island Marsh

Ribbed Marsh Mussel

Geukensia demissa

6 of 9

Linear Regression

Mummichog

Fundulus heteroclitus

Key Points:

​

  • y=mx+b
  • R2 - the square of the correlation coefficient
  • Benchwork assumption R>.9
  • Fieldwork or social science assumption R>.7

R > .9 or R < -.9

Very high positive (negative) correlation

R > .7 or R < -.7

High positive (negative) correlation

R > .5 or R < -.5

Moderate positive (negative) correlation

7 of 9

t-test (Independent Samples)

Key Points:

​

  • If p>.05 then there is no statistically significant difference between groups
  • If p<.05 then there is a statistically significant difference between groups

Mosquito

Family Culicidae

8 of 9

Interpreting Bell Curves (simplified)

9 of 9

Interpreting Bell Curves (simplified)

Equal variances NOT assumed

p<.05

Equal variances assumed

p>.05

F test for equal variances (Homogeneity of Variance) - placement on the number line is irrelevant

OR

t test for means testing - placement on the number line is important