Question

CHALLENGE QUESTION

Nematode worms of the species C. elegans come in two sexes: there are hermaphrodites and males (no females). Fertilized eggs incubated at lower temperatures usually develop into hermaphrodites, while fertilized eggs incubated at higher temperatures can develop into hermaphrodites or into males.

Mark and Christina took a genetics lab where they counted the number of male and hermpahrodite worms that developed from two separate groups of fertilized eggs (Group 1 and Group 2). Group 1 eggs were incubated at 15 degrees C and Group 2 at 25 degrees C. When it was time to analyze their data, Mark and Christina got into a heated argument about which statistical test to use.

Mark insisted that they should analyse their data with a t-test because:

- they are testing a continuous variable (temperature);

- their experiment is a separate treatments experiment;

- there could be any number of males and hermaphrodites in each group.

Christina argued that they should use a chi-square test because:

- in their genetics course they has always used chi-square tests for everything they did;

- in their experiment they are dealing with absolute frequencies subdivided into (two) categories;

- their response of interest is categorical/non-continuous response (developing into a male vs. developing into a hermaphrodite) and not numerical

So, which test should Mark and Christina actually use, and why? (Please select all the answers that apply).

1. They should use a t-test because they are testing a continuous variable (temperature).

2. They should use a t-test because their experiment is a separate treatments experiment.

3. They should use a t-test because there could be any number of males and hermaphrodites in each group.

4. They should use a Chi-square test because in their genetics course they has always used chi-square tests for everything they did.  

5. They should use a Chi-square test because in their experiment they are dealing with absolute frequencies subdivided into (two) categories.

6. They should use a Chi-square test because their response of interest is categorical/non-continuous response (developing into a male vs. developing into a hermaphrodite) and not not numerical.

I think the answer would be chi square but I'm not sure. Thanks!

Answer 1

They should use a Chi-square test because their response of interest is categorical/non-continuous response (developing into a male vs. developing into a hermaphrodite) and not not numerical.

Chi-square test would be the best option here because they have to check whether temperature (variable 1) and development of eggs into hermaphrodites or males are dependent or independent of each other.

A T-Test would not be the choice of the test here, because we are not comparing the results which are two-sided (tails) and we cannot compare the mean of the results from these two groups.