3-4-12. Hypothesis Testing

Once we set up our null and alternative hypotheses, we need used data to figure out which hypothesis we actually think is more likely to be true. And there are two ways we might approach choosing one of these hypotheses. One is the approach that we saw using confidence intervals, where we simulate the sampling … Read more

Now that you’ve had a chance to gain some familiarity with setting up hypothesis test, let’s look at some of the most common tests that are done in practice. One common test is to test the mean or proportion of a population being equal to some value. For example, in finance, we could have the … Read more

Hopefully, you’re starting to feel more comfortable identifying hypotheses and the types of errors that you can make. In this video, we’ll go through one more example to iterate on these ideas. This example is one that really helped me connect all the dots of hypotheses testing and type I and type II errors. I … Read more

From the quizzes, you’ve now been introduced to two types of errors. Type one and type two errors. Correctly setting up the null and alternative hypotheses is important for exactly this reason. They define the importance of the errors that we’re making. In the previous example, a type one error is one where we decided … Read more

Now that you’ve had some practice with setting up hypotheses, you might be asking yourself, why does all this null and alternative stuff really matter? Well, it actually matters a lot. Let’s consider again the judicial example from before. There are four potential reality decision combos that could be made. In order to look at … Read more

In the previous video, you were introduced to the terms of the null and alternative hypotheses and came up with a few guidelines for setting these up as shown here. You also saw how the claim of innocent until proven guilty relates to the null and alternative hypotheses. In this video, I want to provide … Read more

When performing hypothesis testing, the first thing we need to do is translate a question into two competing hypotheses. One of these hypotheses is called the Null, and it’s associated with the symbol, and the other is called the Alternative and it’s commonly notated in this way. Setting up these hypotheses can be a bit … Read more

Academic and industry professionals have questions about, well, just about everything. As data analysts, we try to help them answer these questions. But first, we need to translate the questions into what are known as hypothesis. Then, we need to be able to collect data to justify which hypothesis are likely to be true. As … Read more

You now have learned how to set up a null and alternative hypothesis. You have determined type one and type two errors. And you can calculate which hypothesis you should choose based on error threshold. You also have learned about some of the dangers of the conclusions that you might make when having a really … Read more

So you saw earlier that hypothesis tests and confidence intervals are pretty similar. But you might be surprised just how similar these two techniques are. It turns out that if you are to do a hypothesis test that contains a not equal in alternative hypothesis, your conclusions are identical to a confidence interval, as long … Read more

You’ve learned about the types of errors that are possible on hypothesis testing. And you’ve learned about how we can create a threshold for how often we allow these to happen. But consider if we run 20 of the exact same types of hypothesis tests. Even if the null is actually true, if we have … Read more

When conducting a hypothesis test, there are really two components that you have control over. First, you should be asking yourself, is my sample representative of my population of interest? Are there ways to assure that everyone in my population is accurately represented in my sample? If your sample isn’t representative, then your conclusions are … Read more

When making a decision about whether you’re choosing the null or alternative hypothesis, you might see certain professionals, specifically statisticians, cringe, with concluding remarks like, “So based on the data, we accepted the null hypothesis to be true,” or, “Based on the data, we accepted the alternative.” Remember, when setting up our null and alternative … Read more

In the last video, we saw that a p-value is the probability of obtaining our data or more extreme values from the null hypothesis. So how does this connect to making decisions and the types of errors that we can make? If the p-value is really small, this suggests it’s less likely to observe our … Read more

In the last programming video, you saw how we could simulate draws from the null hypothesis, and that if our statistic was in the bulk of the distribution, this suggested that the statistic was likely from that null. However, if the statistic was farther out from the bulk of the distribution, this suggested that the … Read more

In the previous video, you saw two methods for how we might choose between competing hypotheses. In the second method, we ask the question if the null hypothesis is true, what is the probability of obtaining the statistic we observed in our data or one more extreme in favor of the alternative hypothesis? This probability … Read more

Consider the same example as earlier, where we asked if the mean height for all coffee drinkers was greater than 70 inches. We could again set up a null and alternative hypotheses like these. A second approach that is commonly done for making decisions in hypothesis testing is the following, we assume that the null … Read more

I have attached the data for this example in the Resources tab. And you also can find it in the works pace on the next concept of our work through. This is the same coffee data you saw in the previous lesson on confidence intervals. So you might already have it saved on your local … Read more

Welcome to this lesson on hypothesis testing. Hypothesis testing is one of my favorite topics, because it really does bring an art to the way that we think about statistics. This can bring a lot of complexity. And you may need to watch videos in the section multiple times, or practice the quizzes multiple times. … Read more