Welcome to GrabNGoInfo! What is a p-value is one of the most commonly asked questions in a data science interview. In this tutorial, you will learn:

- What are the general strategies for answering a definition type of interview question?
- How to explain p-value to a technical audience?
- How to explain p-value to a non-technical audience?

**Resources for this post:**

- More video tutorials on statistics
- More blog posts on statistics
- Click here for the slides
- If you prefer the video version of the tutorial, watch the video below on YouTube.

Letâ€™s get started!

### General Strategies

Firstly, letâ€™s talk about the general strategy for answering a definition type of interview question.

- The first strategy is
**concise.**In a data science technical interview, there is usually a list of questions that an interviewer needs to go through with the interviewee. Therefore, time management in an interview is important. We need to keep the answers clear and concise for the definition-type questions and save time for the exploration-type questions. - The second strategy is to talk about the
**key values**for the concept. This includes the value range, how to interpret a large or a small value, and the important threshold value. - The third strategy is to prepare
**simple examples**to support the definition. You do not need to talk about the examples when answering the questions, but keep them in the back of your pocket in case there are follow-up questions asking you to provide an example.

### How to explain p-value to a technical audience?

**Concise Definition**: P-value measures the probability of obtaining results that are at least as extreme as the observed results if the null hypothesis is true.

**Key Values**: P-value ranges from 0 to 1.

- A high p-value means that there is a high chance of obtaining the results as extreme as observed if the null hypothesis is true, so we fail to reject the null hypothesis.
- A low p-value means that there is a low chance of obtaining the results as extreme as observed if the null hypothesis is true, so we reject the null hypothesis.
- The typical threshold for p-value is 0.05, meaning that if the probability of obtaining the results as extreme as observed given the true null hypothesis is less than 0.05, we reject the null hypothesis.

**Simple Example**: To test if a coin is a fair coin, the coin was thrown 100 times. Out of 100 times, 30 times are heads and 70 times are tails. In this case,

- The observed results are 30 heads and 70 tails.
- The results that are at least as extreme as the observed results are headsâ‰¤30 or tailsâ‰¤30.
- P-value measures the probability of getting at least as extreme as the observed results (P(headsâ‰¤30)+P(tailsâ‰¤30)) if the coin is a fair coin.

### How to explain p-value to a non-technical audience?

**Concise Definition**: P-value measures how confident we are about the assumption given the observed data.

**Key Values**: P-value ranges from 0 to 1.

- A high p-value means that we are very confident about the assumption based on the observed data, so we believe the assumption is true.
- A low p-value means that we are not confident about the assumption based on the observed data, so we believe the assumption is not true.
- The typical threshold used for p-value is 0.05, meaning that we believe the assumption is not true if p-value is less than or equal to 0.05.

**Simple Example**: If we assume a coin is a fair coin and the experiment of throwing it multiple times gives us a p-value of 0.01. Because 0.01 is less than 0.05, we are not confident that this coin is fair, so the conclusion is that the coin is unfair.

### Summary

In this tutorial, you learned:

- What are the general strategies for answering a definition type of interview question?
- How to explain p-value to a technical audience?
- How to explain p-value to a non-technical audience?

For more information about data science and machine learning, please check out myÂ YouTube channelÂ andÂ Medium PageÂ or follow me onÂ LinkedIn.

### Recommended Tutorials

- GrabNGoInfo Machine Learning Tutorials Inventory
- Power Analysis For Sample Size Using Python
- Time Series Forecasting Of Bitcoin Prices Using Prophet
- Multivariate Time Series Forecasting with Seasonality and Holiday Effect Using Prophet in Python
- Four Oversampling And Under-Sampling Methods For Imbalanced Classification Using Python
- Recommendation System: User-Based Collaborative Filtering
- Sentiment Analysis Without Modeling: TextBlob vs. VADER vs. Flair