The chi-square distribution is a family of right-skewed density curves with only positive values, defined by degrees of freedom, that models the chi-square statistic and gives you p-values for tests on categorical data like goodness of fit.
The chi-square distribution is the reference curve for chi-square tests. When you compute a chi-square statistic (which measures how far observed counts sit from expected counts, relative to those expected counts), you compare it to a chi-square distribution to get a p-value. It's not one curve but a whole family of curves, and degrees of freedom pick which family member you use.
The shape rules are the part the AP exam loves. Chi-square distributions take only positive values (the statistic is a sum of squared terms, so it can never be negative) and they are skewed right. As degrees of freedom increase, the skew softens, the curve shifts right, and the distribution looks more and more like a normal curve. Think of low degrees of freedom as a curve smashed against zero, and high degrees of freedom as that curve relaxing into a near-normal bell.
This term lives in Unit 8 (Inference for Categorical Data: Chi-Square), specifically Topic 8.2, and it's the direct target of learning objective 8.2.A, which asks you to describe chi-square distributions. The distribution is the engine behind every chi-square test in the unit, including the goodness-of-fit test (8.2.C). It also explains the conditions you check before testing. The large counts condition (all expected counts greater than 5, per 8.2.E) exists because the chi-square distribution only accurately models your statistic when counts are big enough. So when an FRQ asks you to verify conditions, you're really verifying that this distribution is a fair approximation.
Keep studying AP Statistics Unit 8
Chi-Square Statistic (Unit 8)
The statistic is the number you calculate from your data; the distribution is the curve you compare it to. The statistic sums (observed minus expected) squared over expected, and the chi-square distribution tells you whether that sum is surprisingly large.
Degrees of Freedom (Unit 8)
Degrees of freedom (categories minus 1 for goodness of fit) select which chi-square curve you use. More degrees of freedom means less skew, a larger center, and a bigger critical value at the same significance level.
Expected Count (Unit 8)
Expected counts (sample size times null proportion) are what the null hypothesis predicts. The chi-square distribution only behaves correctly when all expected counts are above 5, which is exactly why the large counts condition exists.
Goodness of Fit Test (Unit 8)
The goodness-of-fit test is the first place you actually use this distribution. You test whether one categorical variable's observed distribution matches a claimed set of proportions, and the p-value comes from a chi-square curve.
Multiple-choice questions hit the shape facts hard. Expect stems comparing a chi-square distribution with 5 degrees of freedom to one with 20, asking what happens as degrees of freedom increase (the curve becomes less skewed and approaches a normal distribution), why that happens, and how the critical value at α = 0.05 changes as df rises from 1 to 30 (it gets larger). On FRQs, the distribution shows up indirectly. When you run a chi-square test, you name the test, check that all expected counts exceed 5 so the chi-square approximation is valid, state degrees of freedom, and use the distribution to find your p-value. You won't be asked to derive the curve, but you need to describe it and know when it applies.
The chi-square statistic is a single number computed from your sample, the sum of (observed − expected)²/expected across categories. The chi-square distribution is the theoretical curve that statistic follows when the null hypothesis is true. You calculate the statistic, then locate it on the distribution to find the p-value. Mixing these up leads to answers like 'the statistic is skewed right,' which describes the distribution, not the number.
Chi-square distributions take only positive values and are skewed right, because the chi-square statistic is built from squared differences.
Degrees of freedom determine which chi-square curve you use, and as degrees of freedom increase, the skew decreases and the curve approaches a normal distribution.
As degrees of freedom increase, the critical value for a given significance level (like α = 0.05) also increases.
The chi-square statistic measures the distance between observed and expected counts relative to expected counts, and the distribution converts that distance into a p-value.
The large counts condition (all expected counts greater than 5) ensures the chi-square distribution is an accurate model for your test statistic.
Expected counts come from the null hypothesis and equal sample size times the null proportion for each category.
It's a family of right-skewed density curves with only positive values, indexed by degrees of freedom. In Unit 8 you use it to find p-values for tests on categorical data, like the goodness-of-fit test in Topic 8.2.
Not exactly, but it gets close. Every chi-square distribution is skewed right, and the skew becomes less pronounced as degrees of freedom increase, so a chi-square curve with 30 degrees of freedom looks approximately normal while one with 1 or 2 degrees of freedom is piled up near zero.
The statistic is the number you calculate from data, the sum of (observed − expected)²/expected. The distribution is the theoretical curve that statistic follows under the null hypothesis, and it's where your p-value comes from.
More degrees of freedom means the statistic sums more squared terms, which pushes the distribution's center to the right and smooths it toward a normal shape. AP multiple-choice questions ask this directly, so know that higher df means less skew and a curve approaching normal.
No. The chi-square statistic squares every observed-minus-expected difference, so it's always zero or positive, which is exactly why the distribution only has positive values. A negative chi-square value on your calculator means you made an arithmetic error.