LAB ____: THE CHI-SQUARE TEST
LAB ____: THE CHI-SQUARE TEST
LAB ____: THE CHI-SQUARE TEST
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
Name __________________________________ Period _________<br />
AP Biology Date ______________________<br />
1. What is the chi-square test used for?<br />
<strong>LAB</strong> ____: <strong>THE</strong> <strong>CHI</strong>-<strong>SQUARE</strong> <strong>TEST</strong><br />
The chi-square test is a statistical test that makes a comparison between the data<br />
collected in an experiment versus the data you expected to find. The chi-square test is a<br />
way to evaluate this variability to get an idea if the difference between real and expected<br />
results are due to normal random chance, or if there is some other factor involved (like<br />
an unbalanced coin or a different inheritance scheme than the one you predicted).<br />
2. Why is probability important in genetics?<br />
Most eukaryotic organisms are diploid. In preparation for sexual reproduction, the two<br />
copies of each chromosome first separate are randomly sorted into the gametes during<br />
meiosis. When gametes from two different parents combine in fertilization, new<br />
combinations of alleles are created. Thus chance plays a major role in determining which<br />
alleles, and therefore which combinations of traits end up in each new individual, so the<br />
inheritance of characteristics is the result of random chance and can be calculated by<br />
using the laws of probability.<br />
3. Briefly describe how the chi-square analysis may be used in genetics.<br />
In classical genetics research where you are trying to determine the inheritance pattern<br />
of a phenotype, you establish your predicted genetic explanation and the expected<br />
phenotype ratios in the offspring (from your Punnett square) as your null hypothesis.<br />
After you make your crosses, you would then compare your observed results (phenotype<br />
counts) against your expected results and complete a chi-square analysis. If the p value is<br />
determined to be greater than .05 then you would accept your null hypothesis (differences<br />
are due to random chance alone) and your genetic explanation for this trait is supported.<br />
If the p value is determined to be .05 or less then you would reject your null hypothesis—<br />
random chance alone can only explain this level of difference fewer than 1 time out of<br />
every 20 times—and your genetic explanation for this trait is unsupported. You therefore<br />
have to consider alternative inheritance patterns for the trait you are researching.<br />
4. Suppose you were to obtain a chi-square value of 7.82 or greater in your data analysis (with<br />
2 degrees of freedom). What would this indicate?<br />
Since 7.82 is between 5.99 (p=.05) and 9.21 (p=.01), I would reject my null hypothesis. If<br />
it were in coin flips, then I would have to say that there was something wrong with my<br />
coin. If it were in a genetics experiment, then I would have to say that my inheritance<br />
pattern model is not supported and I would have to suggest another model.<br />
5. Suppose you were to obtain a chi-square value of 4.60 or lower in your data analysis (with 2<br />
degrees of freedom). What would this indicate?<br />
Since 4.60 is at p=.0.10, I would accept my null hypothesis. If it were in coin flips, then I<br />
would be allowed to say that my coin was a “fair” coin. If it were in a genetics<br />
experiment, then I would be allowed to say that my inheritance pattern model is<br />
supported.<br />
1 of 5<br />
Adapted by Kim B. Foglia • www.ExploreBiology.com • 2005-2006
Name __________________________________ AP Biology<br />
6. A heterozygous white-fruited squash plant is crossed with a yellow-fruited plant, yielding 200<br />
seeds. Of these, 110 produce white-fruited plants while only 90 produce yellow-fruited<br />
plants. Are these results statistically significant? Explain using chi-square analysis.<br />
Expected: 1 yellow : 1 white ratio<br />
w w<br />
W Ww Ww<br />
w ww ww<br />
A B C D E<br />
Obs Exp Obs - Exp (Obs – Exp) 2 (Obs – Exp) 2<br />
Exp<br />
white-fruit 110 100 10 100 1<br />
yellow-fruit 90 100 -10 100 1<br />
2 of 5<br />
Adapted by Kim B. Foglia • www.ExploreBiology.com • 2005-2006<br />
X 2 Total 2<br />
Degrees of Freedom 1<br />
At x 2 = 2 (df 1), you would get a p value > 0.30 and therefore you would accept your null<br />
hypothesis and you could say that your proposed inheritance pattern was supported.<br />
7. What if there were 2000 seeds and 1100 produced white-fruited plants & 900 yellow-fruited?<br />
Are these results statistically significant? Explain using chi-square analysis.<br />
Expected: 1 yellow : 1 white ratio<br />
w w<br />
W Ww Ww<br />
w ww ww<br />
A B C D E<br />
Obs Exp Obs - Exp (Obs – Exp) 2 (Obs – Exp) 2<br />
Exp<br />
white-fruit 1100 1000 100 10,000 10<br />
yellow-fruit 900 1000 -100 10,000 10<br />
X 2 Total 20<br />
Degrees of Freedom 1<br />
At x 2 = 20 (df 1), you would get a p value
Name __________________________________ AP Biology<br />
8. TEAM DATA: What was your hypothesis (expected values) for your individual team coin<br />
toss?<br />
Exp<br />
H/H 25<br />
H/T 50<br />
T/T 25<br />
What was your calculated chi-square value for your individual team data? _______________<br />
Answers will vary<br />
What p value does this chi-square correspond to? __________________________________<br />
Answers will vary<br />
Was your hypothesis supported by your results? Explain using your chi-square analysis.<br />
___________________________________________________________________________<br />
Answers will vary<br />
___________________________________________________________________________<br />
___________________________________________________________________________<br />
___________________________________________________________________________<br />
9. CLASS DATA: What was your hypothesis (expected values) for the class coin toss?<br />
___________________________________________________________________________<br />
Answers will vary<br />
___________________________________________________________________________<br />
What was your calculated chi-square value for the class data? ________________________<br />
Answers will vary<br />
What p value does this chi-square correspond to? __________________________________<br />
Answers will vary<br />
Was your hypothesis supported by the results? Explain using your chi-square analysis.<br />
___________________________________________________________________________<br />
Answers will vary<br />
___________________________________________________________________________<br />
3 of 5<br />
Adapted by Kim B. Foglia • www.ExploreBiology.com • 2005-2006
Name __________________________________ AP Biology<br />
10. When scientists design research studies they purposely choose large sample sizes. Work<br />
through these scenarios to see why:<br />
a. Just as in your experiment, you flipped 2 coins, but you only did it 10 times. You<br />
collected these data below. Use the chart to calculate the chi-square value:<br />
Obs Exp Obs - Exp (Obs – Exp) 2 (Obs – Exp) 2<br />
Exp<br />
H/H 1 2.5 -1.5 2.25 .9<br />
H/T 8 5 3 9 1.8<br />
T/T 1 2.5 -1.5 2.25 .9<br />
4 of 5<br />
Adapted by Kim B. Foglia • www.ExploreBiology.com • 2005-2006<br />
X 2 Total 3.6<br />
Would you accept or reject the null hypothesis? Explain using your chi-square analysis.<br />
With X 2 =3.6 and df=2, would accept null hypothesis with p>0.10<br />
b. Now you flipped your 2 coins again, but you did it 100 times. You collected these data<br />
below. Use the chart to calculate the chi-square value:<br />
Obs Exp Obs - Exp (Obs – Exp) 2 (Obs – Exp) 2<br />
Exp<br />
H/H 10 25 -15 225 9<br />
H/T 80 50 30 900 18<br />
T/T 10 25 -15 225 9<br />
X 2 Total 36<br />
Would you accept or reject the null hypothesis? Explain using your chi-square analysis.<br />
With X 2 =36 and df=2, would reject null hypothesis with p
Name __________________________________ AP Biology<br />
d. Now, using your understanding of the chi-square test, explain why scientists purposely<br />
choose large sample sizes when they design research studies.<br />
With a small sample size, random chance is “allowed” to play a larger role in the<br />
variation of the data (the difference between observed and expected). In other words, the<br />
probability of getting a ratio of 1:8:1 in 10 flips by random chance is high enough that<br />
the chi-square analysis would allow you to conclude that there is nothing wrong with<br />
your coins. But with a larger sample size, random chance is not “allowed” to play a such<br />
a role. The probability of getting a ratio of 1:8:1 in 100 or 1000 flips by random chance<br />
is so low that the chi-square analysis would force you to conclude that there is something<br />
wrong with your model—you did not use fair coins or you are not a fair flipper. Another<br />
way to look a this is if they are testing the efficacy of a drug, do you want them to test it<br />
on 10 subjects or 10,000 subjects?<br />
5 of 5<br />
Adapted by Kim B. Foglia • www.ExploreBiology.com • 2005-2006