0
$\begingroup$

I have some data which is positively skewed $(0.9)$ and I would like to be able to find the probability of a given value. With normally distributed data I know that $68\%$ of data is within one standard deviation etc. and you can use a probability table to work out a value's probability. However, is there a way to do this with skewed data? Any help would be appreciated, thanks in advance!

  • 0
    do you have an actual distribution or are you estimating a distribution from data? – 2017-02-05
  • 0
    I have worked out the skewness, mean, standard deviation, IQR etc. from my data? – 2017-02-05
  • 0
    The values you can look up in a table are worked out as with any distribution, i.e. $P(X \leq x) = \int^x_{-\infty} f(u)du$ where $f(u)$ is the density of $X$. So in your case you would have to start by fitting a skewed distribution, like the beta distribution. Alternatively, do some kernel estimation. – 2017-02-05
  • 0
    So I need to plot my data on a graph to get the equation and then I can work out the probabilities? – 2017-02-05
  • 0
    The distribution you want to fit to your sample has to based on which assumptions you make about the nature of the data. Maybe add more info on what kind of data this is and what you are trying to do. But that's becoming more of a statistics problem than math. – 2017-02-05

0 Answers 0