Resources
An ecologist tests whether the spatial arrangement of creosote bushes (Larrea tridentata) across a desert study plot differs significantly from a Poisson random distribution. The plot is partitioned into contiguous grid quadrats. The observed counts of quadrats containing bushes are summarized as follows: : quadrats; : quadrats; : quadrats; : quadrats. The total number of observed bushes is . (a) Determine the mean density in . (b) Using the Poisson probability mass function , calculate the expected frequencies for , and (where ). (c) Compute the goodness-of-fit test statistic , and state whether you reject or fail to reject the null hypothesis of a random spatial distribution at (critical value , with due to estimating ).
Answer and steps
Final answer
(a) ; (b) , , , ; (c) ; fail to reject the null hypothesis of a random spatial distribution
Steps
Part (a): Determine the Mean Density ()
The parameter represents the average number of creosote bushes per quadrat. Dividing the total number of observed individuals by the total number of sampled quadrats yields:
Part (b): Calculate Expected Frequencies ()
Under the null hypothesis (), individuals are distributed randomly following a Poisson distribution with parameter :
First, evaluate the base exponential term:
For a sample size of quadrats, the expected frequency for each count category is :
- For :
- For :
- For :
- For : Because all probabilities must sum to :
Part (c): Goodness-of-Fit Test and Hypothesis Conclusion
Calculate the chi-square () test statistic using the formula:
Compute each component from the observed () and expected () frequencies:
- :
- :
- :
- :
Summing the components:
Hypothesis Decision:
- Degrees of freedom: (loss of degree of freedom for the fixed total quadrats and for estimating the mean parameter ).
- Critical value: .
Because , the test statistic does not exceed the critical threshold (). Therefore, we fail to reject the null hypothesis of a random spatial distribution. There is no statistically significant evidence that the distribution of creosote bushes differs from a Poisson random distribution.
Typical errors
01 The student calculates λ by dividing the total number of bushes by the total area in square meters instead of by the number of quadrats, leading to an incorrect parameter value and flawed expected frequencies.
Sample Student Work
Reviewed by ACEpath TeacherStep-by-Step Review
Walk Through It with Your Teacher- Step 1Correct
You divided the total count of bushes () by the total area () instead of the total number of quadrats (). Because the random variable counts bushes per quadrat, the mean parameter must be expressed per quadrat.
- Step 2Correct
Your formulas for evaluating the Poisson probabilities and expected frequencies were mathematically sound, but your numerical values were thrown off by using instead of .
- Step 3Correct
You applied the chi-square goodness-of-fit formula correctly to your expected values, but the resulting statistic was inflated due to the propagated errors in .
- Step 4Correct
You correctly compared your computed test statistic against the critical value, but because the test statistic was artificially inflated, your final ecological conclusion was reversed.
WHAT YOU DID WELL
Compliments from Your TeacherYou correctly identified and applied the Poisson probability mass function .
You used the complement rule correctly to calculate the probability and expected count for the open-ended class .
You correctly structured the goodness-of-fit statistic and compared it against the given critical value.
THE ERROR
Where Your ACEpath Teacher Spotted the MistakeYou divided by the total area of rather than the sample size of quadrats.
WHY IT HAPPENED
ACEpath Teacher's DiagnosisYou confused spatial density per unit area (bushes per square meter) with the mean discrete count per sampling unit (bushes per quadrat) required by the Poisson distribution.
REASONING PATH
Your ACEpath teacher’s suggested approach- Define the Sampling Unit
In quadrat sampling, the Poisson parameter represents the mean count per quadrat, given by .
Always check the denominator of : divide by the count of quadrats (), never the area in square meters, unless each quadrat is exactly .
CORRECTED PATH
ACEpath Teacher's DemonstrationRecalculate : .
Recalculate expected counts: , , , and .
Compute : .
(a) ; (b) , , , ; (c) ; fail to reject the null hypothesis of a random spatial distribution.
02 The student calculates the expected frequency for the tail class by evaluating P(3) rather than computing the cumulative tail probability P(k ≥ 3).
Sample Student Work
Reviewed by ACEpath TeacherStep-by-Step Review
Walk Through It with Your Teacher- Step 1
- Step 2
- Step 3Correct
You evaluated only the point probability rather than the open-ended tail probability . Because a Poisson distribution is supported on all non-negative integers up to infinity, all categories must sum to . Using the complement rule yields , giving instead of .
- Step 4Correct
You correctly structured the goodness-of-fit formula , but the fourth term was affected by using instead of .
- Step 5Correct
Your comparison logic and conclusion are valid, but they rely on the affected test statistic value rather than the correct value .
WHAT YOU DID WELL
Compliments from Your TeacherYou correctly calculated the sample mean density .
You accurately evaluated Poisson point probabilities and expected frequencies for , and .
You correctly applied the chi-square goodness-of-fit formula and correctly compared the test statistic against the critical value .
THE ERROR
Where Your ACEpath Teacher Spotted the Mistakeno
WHY IT HAPPENED
ACEpath Teacher's DiagnosisYou treated an open-ended terminal class () as a single point mass (), omitting the probabilities for and causing the expected frequencies across all categories to sum to less than the total sample size.
REASONING PATH
Your ACEpath teacher’s suggested approach- Exhaustive Category Sum Rule
In a chi-square goodness-of-fit test, the sum of all expected frequencies must equal the total number of observations: .
Always verify that and . If your final category is open-ended (e.g., ), calculate its probability as .
CORRECTED PATH
ACEpath Teacher's DemonstrationApply the complement rule to capture the full tail: .
Compute the expected count: quadrats.
Recompute the terminal chi-square component: (or ).
Sum all four components: .
; since , fail to reject the null hypothesis of a Poisson random spatial distribution.
03 The student divides squared differences by observed counts instead of expected frequencies in the chi-square formula denominator.
Sample Student Work
Reviewed by ACEpath TeacherStep-by-Step Review
Walk Through It with Your Teacher- Step 1
- Step 2
- Step 3Correct
In the Pearson chi-square formula , each squared deviation must be scaled by the expected frequency , not the observed count .
- Step 4Correct
Your comparison and final hypothesis decision are logically sound, but your calculation was carried out from an incorrect formula denominator.
WHAT YOU DID WELL
Compliments from Your TeacherYou correctly calculated the sample mean density across all quadrats.
You correctly applied the Poisson probability mass function to compute expected frequencies for all categories, including the tail category .
THE ERROR
Where Your ACEpath Teacher Spotted the MistakeYou accidentally divided each squared deviation by the observed frequency instead of the expected frequency .
WHY IT HAPPENED
ACEpath Teacher's DiagnosisConfusing the roles of observed and expected counts in the chi-square test statistic. Under the null hypothesis, the variance of a Poisson or multinomial count category is proportional to its expected value , so standardizing the squared discrepancy requires dividing by , not .
REASONING PATH
Your ACEpath teacher’s suggested approach- Standardizing Deviations
Pearson's chi-square test measures normalized deviation from null expectations: .
Always write down the general formula and check that every denominator matches your calculated expected value for that bin.
CORRECTED PATH
ACEpath Teacher's DemonstrationSubstitute expected frequencies into the denominator:
Compute each term:
Compare with critical value: , so fail to reject .
(a) ; (b) , , , ; (c) ; fail to reject the null hypothesis of a random spatial distribution
Lock it in
When testing quadrat counts against a Poisson distribution, what is the denominator used to calculate the mean parameter ?
The total number of sampled quadrats (). represents the expected number of individuals per sampling unit (quadrat), not per square meter.
Similar question
An ecologist samples barnacles using quadrats of size placed across a rocky shore. A total of barnacles are counted. (a) Determine the Poisson parameter in . (b) Calculate the expected number of quadrats containing exactly zero barnacles, .
Answer
(a) . (b) . Expected quadrats quadrats.
A forest survey divides a plot into quadrats of each. A total of saplings are recorded. (a) Find in . (b) Calculate the expected frequency of quadrats with exactly sapling () and exactly saplings ().
Answer
(a) . (b) . .
A field survey uses quadrats to count dandelion plants, finding a total of plants. The observed counts are: plants in quadrats, plant in quadrats, and plants in quadrats. Using , compute the expected counts and the chi-square statistic .
Answer
Expected probabilities: , , . Expected frequencies: , , . Chi-square statistic: .