What is a Q-Q plot and how can it help test for common distributions?

  • Thread starter Tosh5457
  • Start date
  • Tags
    Plot
In summary: You could then create a scatter plot with one set of quantiles on one axis and the other set on the other axis. In summary, a Q-Q plot is a way to test if two sets of data come from the same distribution. It involves plotting quantiles of the two data sets against each other and comparing them to a 45 degree line. It can be useful for testing distributional aspects and can be made by gathering estimated quantiles from the data sets and creating a scatter plot.
  • #1
Tosh5457
134
28
I've read the article in wikipedia but I can't understand it. Can someone point me to a good explanation of what a Q-Q plot shows and how to make one, or explain to me please?
 
Physics news on Phys.org
  • #2
Tosh5457 said:
I've read the article in wikipedia but I can't understand it. Can someone point me to a good explanation of what a Q-Q plot shows and how to make one, or explain to me please?

You should supply a link to the article.
 
  • #4
Oops, answered the wrong question
 
  • #5
The q-q plot is a common way to test if two sets of data come from the same distrbution. Typically, you'll plot a 45 degree line and with the two quantiles versus each other and if they come from the same distribution, then they should line up with the 45 degree line. The more the points depart from the line the more evidence you have that they are not common distributions. A q-q plot has the advantage that you can test two sample sizes of unequal size against each other, and at the same time test some distributional aspects (ie shifts, symmetry, outliners.)

For as how to make one, I suppose the first step would be to gather estimated quantiles from the two data sets.
 

1. What is a Q-Q plot?

A Q-Q plot, short for quantile-quantile plot, is a graphical tool used to compare the distribution of two datasets. It plots the quantiles of one dataset against the quantiles of another dataset, allowing for visual assessment of how similar or different their distributions are.

2. How do you interpret a Q-Q plot?

The closer the points on a Q-Q plot are to the diagonal line, the more similar the distributions of the two datasets are. If the points fall along the diagonal line, it indicates that the two datasets have the same distribution. Deviations from the diagonal line suggest differences in the shape, spread, or location of the distributions.

3. What does it mean if a Q-Q plot is skewed?

A skewed Q-Q plot indicates that the two datasets being compared have different distributions. This means that the data may not follow a normal distribution and may require different statistical methods for analysis.

4. How do you create a Q-Q plot?

To create a Q-Q plot, you first need to have two datasets with numerical values. Then, you can use a statistical software or programming language to plot the quantiles of one dataset against the quantiles of the other dataset. Alternatively, you can create a Q-Q plot by hand using graph paper and calculating the quantiles manually.

5. Can a Q-Q plot be used for non-numerical data?

No, a Q-Q plot is specifically designed for comparing the distributions of numerical data. It cannot be used for non-numerical data, such as categorical or ordinal data. In these cases, other graphical or statistical methods should be used to compare the distributions.

Similar threads

I Translate compound proposition p → q (implication) to p↓q question
    • Set Theory, Logic, Probability, Statistics
Replies
4
Views
866
I How to plot a scaled lognormal function
    • Set Theory, Logic, Probability, Statistics
Replies
12
Views
1K
A How to perform a fit with correlated variables
    • Set Theory, Logic, Probability, Statistics
Replies
16
Views
2K
I Calculate limits as distributions
    • Calculus
Replies
4
Views
1K
B Understanding the Interpretation of Z-Scores: Insights and Examples
    • Set Theory, Logic, Probability, Statistics
Replies
5
Views
442
Mathematica Understanding the Code: How to Extract Data from a Plot in Mathematica
    • MATLAB, Maple, Mathematica, LaTeX
Replies
1
Views
2K
Practical Info on Ultra-High Q Lasers?
    • Electrical Engineering
Replies
5
Views
686
A Bin Width for Overlay Plots of Data + Estimation
    • Set Theory, Logic, Probability, Statistics
Replies
8
Views
2K
I Why prove Taylor's theorem with p(x)=q(x)−((b−a)^n/(b−x)^n)q(a)?
    • Calculus
Replies
9
Views
825
Heat flow components of Stress/Energy Tensor
    • Advanced Physics Homework Help
Replies
0
Views
545

Hot Threads

Recent Insights

Back
Top