Selecting a Statistical test

gdt_icon_150x.png

How DataClassroom can guide your students in choosing the right test

 

What does it do?

The workflow in DataClassroom guides the student through three stages:

  • Data preparation

  • Visualization (graphing)

  • Statistical tests

Here, we will focus on the critical step from visualization to statistics, where mathematical analysis is used to evaluate a hypotheses relevant to the data.

(See this other post on how we also help with selecting a graph…)

gdt_coffee_square.png
 
t_test_two_groups_example.png

Building on the student’s own intuition

By graphing the data in a way that illustrates the question, for example “Are these groups significantly different?”, the student has already done the groundwork.

When they then click on the Graph-Driven Test button, the tool uses this information.

Based on how the data are graphed, it presents a list of possible statistical tests, and explains:

  • In which situations each would be relevant

  • Which one is recommended for the student’s own graph

 

Demystification

It’s important that there is no “magic” here. The student can clearly see that a T-test has been recommended, because they have data with:

  • A variable (X) with two categorical values that group the data

  • A measurement (Y) that is a sample of numeric values

However, the actual calculation of the T-test results happen at the press of a button.

Note: DataClassroom also has an interactive, animated T-test analysis, which can take the student through the math, if needed.

gdt_t_test_gui.png
 
pslider_question.png
p_slider_0_17.png

Explanation of results

The result of a T-test is the P-value, and this often calls for a long explanation, caveats, etc. We present this as a measurement on an interactive sliding scale, along with additional relevant information.

  • Gives the student a better, more intuitive understanding that the P-value interpretations are not absolutes

  • Allows them to explore the range of possible results, and where theirs lies on the continuum

 
 

Conclusion

We believe passionately that the best learning experience is one that takes place between a teacher and their students.

But also that this experience can be improved by carefully designed tools, that focus not only on getting results but on providing learning opportunities, building student intuition and supporting the teacher by encouraging the students to experiment and learn by doing.

Want to see for yourself?

 
 

Sign up for a webinar or a personal demo, browse our User Guide, or register on the app and try it out for yourself.

 
 
Dan TempleComment