# Gentler approach to Bayesian statistics

I recently started reading "Introduction to Bayesian Statistics" 2nd Edition by Bolstad. I've had an introductory stats class that covered mainly statistical tests and am almost through a class in regression analysis. What other books can I use to supplement my understanding this one?

I've made it through the first 100-125 pages fine. Afterwards the book begins to talk hypothesis testing which is what I'm very excited to cover but there a couple of things throwing me:

• The use of probability density functions in calculations. In other words how to evaluate such equations.
• This whole sentence: "Suppose we use a beta(1,1) prior for pi. Then given y=8, the posterior density is beta(9,3). The posterior probability of the null hypothesis is..." I believe beta(1,1) refers to a PDF where the mean is 1 and the stdev is 1? I don't get how it would change to a beta(9,3) as a posterior density function.

I do get the concept of priors vs posteriors and understand how to apply them using a table manually. I get (I think!) that pi represents the supposed population proportion or probability.

I don't get how to connect this together with data I would run into on a day to day basis and get results.

• The parameter $\pi$ appears from the context to be the population probability of a binomial model. In this case, a beta distribution is the conjugate prior for a binomial likelihood with known $n$ and unknown $\pi$. However, the parameters of the beta distribution are not the mean and standard deviation, as is the case for the normal distribution. Look at the Wikipedia page to see the formula of the mean and variance of a beta random variable in terms of the parameters of the beta distribution. – caburke Aug 2 '13 at 3:11
• Thank you! Conjugate prior is another term that isn't familiar to me. Where can I learn more about that at an introductory level? – Justin Bozonier Aug 2 '13 at 3:22
• You might be interested in a more practical text, have you seen Bayesian Methods for Hackers? (Disclosure - I am a contributing author) Try searching for it (it's opensource and free). – Cam.Davidson.Pilon Aug 2 '13 at 3:26
• @JustinBozonier This link stats.stackexchange.com/questions/66018/… gives some explanation to the different terms people use to describe priors, including conjugate priors. – Sycorax says Reinstate Monica Aug 2 '13 at 12:32
• @Cam.Davidson.Pilon Thanks for that! The updating of beliefs in the charts on this page alone are helping me get more of what the others replying are saying: nbviewer.ipython.org/urls/raw.github.com/CamDavidsonPilon/… – Justin Bozonier Aug 2 '13 at 17:26

The use of probability density functions in calculations. In other words how to evaluate such equations.

I think you're still thinking of this from a frequentist perspective: if you're looking for a point estimate, the posterior won't give it to you. You put PDFs in, you get PDFs out. You can derive point estimates by calculating statistics from your posterior distribution, but I'll get to that in a bit.

I do get the concept of priors vs posteriors and understand how to apply them using a table manually. I get (I think!) that pi represents the supposed population proportion or probability.

$\pi(x)$ is the same thing as $p(x)$: they're both PDFs. $\pi$ is just conventionally used to denote that the particular PDF is a prior density.

I suspect that you don't get priors and posteriors as well as you think you do, so let's back it up to the fundamental underpinning of Bayesian statistics: Subjective Probability.

A Thought Experiment in Subjective Probability

Let's say I present you with a coin and ask you whether or not you think this coin is a fair coin. You've heard a lot of people talk about unfair coins in probability class, but you've never actually seen one in real life, so you respond, "Yeah, sure, I think it's a fair coin." But, the fact that I'm even asking you this question puts you off a little, so although your estimation is that it's fair, you wouldn't really be surprised if it wasn't. Much less surprised than if you found this coin in your pocket change (because you assume that's all real currency, and you don't really trust me right now becaue I'm acting suspicious).

Now, we run a few experiments. After 100 flips, the coin gives back 53 Heads. You're a lot more confident that it's a fair coin, but you're still open to the possibility that it's not. The difference is that now you would be pretty surprised if this coin turned out to have some sort of bias.

How can we represent your prior and posterior beliefs here, specifically, regarding the probability that the coin will show heads (which we will denote $\theta$)? In a frequentist setting, your prior belief--your null hypothesis--is that $\theta = 0.5$. After running the experiment, you're not able to reject the null, and so you continue on with the assumption that yes, the coin is probably fair. But how do we encapsulate the change in your confidence that the coin is fair? After the experiment you are in a position that you would bet that the coin is fair, but before the experiment you would have been trepidatious.

In the Bayesian setting, you encapsulate your confidence in propositions by not treating probabilities as scalar values but as random variables, i.e. functions. Instead of saying $\theta = 0.5$ we say $\theta \sim N(0.5, \sigma^2)$, and thereby encapsulate our confidence in the variance of the PDF. If we set a high variance, we're saying, "I think that the probability is 0.5, but I wouldn't be surprised if the probability I actually observe in the world is far away from this value. I think $\theta= 0.5$, but frankly I'm not really that sure." By setting a low variance, we're saying, "Not only do I believe the probability is 0.5, but I would be very surprised if experimentation provides a value that's not very close to $\theta=0.5$." So, in this example when you start the experiment you have a prior with high variance. After receiving data that corroborates your prior, the mean of the prior stayed the same, but the variance became much narrower. Our confidence that $\theta=0.5$ is much higher after running the experiment than before.

So how do we perform calculations?

We start with PDFs, and we end with PDFs. When you need to report a point estimate, you can calculate statistics like the mean, median or mode of your posterior distribution (depending on your loss function, which I won't get into now. Let's just stick with the mean). If you have a closed form solution for your PDF, it will probably be trivial to determine these values. If the posterior is complicated, you can use procedures like MCMC to sample from your posterior and derive statistics from the sample you drew.

In the example where you have a Beta prior and a Binomial likelihood, the calculation of the posterior reduces to a very clean calculation. Given:

• Prior: $\theta \sim Beta(\alpha, \beta)$
• Likelihood: $X|\theta \sim Binomial(\theta)$

Then the posterior reduces to:

• Posterior: $\theta|X \sim Beta(\alpha + \sum_{i=1}^n x_i,\, \beta + n - \sum_{i=1}^n x_i)$

This will happen any time you have a beta prior and a binomial likelihood, and the reason why should be evident in the calculations provided by DJE. When a particular prior-likelihood model always gives a posterior that has the same kind of distribution as the prior, the relationship between the types of distributions used for the prior and likelihood is called Conjugate. There are many pairs of distributions that have conjugate relationships, and conjugacy is very frequently leveraged by Bayesians to simplify calculations. Given a particular likelihood, you can make your life a lot easier by selecting a conjugate prior (if one exists and you can justify your choice of prior).

I believe beta(1,1) refers to a PDF where the mean is 1 and the stdev is 1?

In the common parameterization of the normal distribution, the two parameters signify the mean and standard deviation of the distribution. But that's just how we parameterize the normal distribution. Other probability distributions are parameterized very differently.

The Beta distribution is usually parameterized as $Beta(\alpha, \beta)$ where $\alpha$ and $\beta$ are called "shape" parameters. The Beta distribution is extremely flexible and takes lots of different forms depending on how these parameters are set. To illustrate how different this parameterization is from your original assumption, here's how you calculate the mean and variance for Beta random variables:

\begin{equation} \begin{split} X &\sim Beta(\alpha, \beta) \\ \operatorname{E}[X] &= \frac{\alpha}{\alpha + \beta} \\ \operatorname{var}[X] &= \frac{\alpha\beta}{(\alpha+\beta)^2(\alpha+\beta+1)} \end{split} \end{equation}

As you can clearly see, the mean and variance are not a part of the parameterization of this distribution, but they have closed form solutions that are simple functions of the input parameters.

I won't go into detail describing the differences in parameterizations of other well known distributions, but I recommend you look up a few. Any basic text, even Wikipedia, should somewhat describe how changing the parameters modifies the distribution. You should also read up on the relationships between the different distributions (for instance, $Beta(1,1)$ is the same thing as $Uniform(0,1)$).

• The key thing your answer gave me was the realization that looking for a single value was where I was getting hung up. Once I started thinking in terms of distributions the Kruschke text and everything else began making much more sense. Thank you! – Justin Bozonier Aug 5 '13 at 13:24

A beta distribution has the form $p(\theta)=\frac{\Gamma(\alpha)\Gamma(\beta)}{\Gamma(\alpha+\beta)}\theta^{\alpha-1}(1-\theta)^{\beta-1}$. A beta(1,1) distribution has parameters $(\alpha, \beta)=(1,1)$. (Unfortunately, this kind of statistical short-hand places a burden on the reader to know how the particular model is parameterized!)

The beta prior with a binomial likelihood (fixed number of trials with binary outcomes and fixed probabilities of success/failure) has the property of conjugacy, which allows the posterior (the product of the prior and the likelihood) to be written in closed form:

\begin{equation} \begin{split} p(\theta|y) &= \frac{p(y|\theta)p(\theta)}{p(y)} \\ ~\\ ~\\ &\propto\frac{\Gamma(\alpha)\Gamma(\beta)}{\Gamma(\alpha+\beta)}\theta^{\alpha-1}(1-\theta)^{\beta-1}*\binom{n}{y}\theta^y(1-\theta)^{n-y} \\ ~\\ ~\\ &\propto\theta^{\alpha-1}(1-\theta)^{\beta-1}*\theta^y(1-\theta)^{n-y} \\ ~\\ &\propto\theta^{\alpha+y-1}(1-\theta)^{\beta+n-y-1} \\ ~\\ &=\frac{\Gamma(\alpha+y-1)\Gamma(\beta+n-y-1)}{\Gamma(\alpha+\beta+n-1)}\theta^{\alpha+y-1}(1-\theta)^{\beta+n-y-1} \end{split} \end{equation}

For the particular example in the text, the author is indicating that a beta(1,1) prior with data n=10 and y=8 produces a beta(1+8,1+2)=beta(9,3) posterior distribution on $\theta$.

This closed-form expression is convenient, but by no means necessary. Multiplying probability densities can be done the same way as multiplying other mathematical expressions; the difficulties arrive since many products of densities are not as easily rewritten as the beta prior/binomial likelihood. Fortunately, this is where computers pick up the slack.

If you are looking for a gentler approach I can highly recommend the book by Kruschke which uses R to explain the core concepts. It is a very practical and hands-on approach into learning Bayesian statistics and on his website you can find all of the codes used.

Someone also recommended the text by Cam.Davidson.Pilon to me, haven't look at it yet but it can be found here.

• Thanks! I actually own the Kruschke book already and just went back to review it and realized it's exactly what I need right now. Thanks for the pointer! – Justin Bozonier Aug 5 '13 at 13:23
• @JustinBozonier I also highly recommend Introduction to the Theory of Statistics (Mood). It provides a relatively high-level of rigor, but only assumes that you know very basic calculus. – Steve P. Aug 8 '13 at 19:29