# Videos for teaching and learning statistics

It delights me that several of my statistics videos have been viewed over half a million times each. As well there is a stream of lovely comments (with the odd weird one) from happy viewers, who have found in the videos an answer to their problems.

In this post I will outline the main videos available on the Statistics Learning Centre YouTube Channel. They already belong to 24,000 playlists and lists of recommended resources in textbooks the world over. We are happy for teachers and learners to continue to link to them. Having them all in one place should make it easier for instructors to decide which ones to use in their courses.

# Philosophy of the videos

Early on in my video production I wrote a series of blog posts about the videos. One was Effective multimedia teaching videos. The videos use graphics and audio to increase understanding and retention, and are mostly aimed at conceptual understanding rather than procedural understanding.

I also wrote a critique of Khan Academy videos, explaining why I felt they should be improved. Not surprisingly this ruffled a few feathers and remains my most commented on post. I would be thrilled if Khan had lifted his game, but I fear this is not the case. The Khan Academy pie chart video still uses an unacceptable example with too many and ordered categories. (January 2018)

Before setting out to make videos about confidence intervals, I critiqued the existing offerings in this post. At the time the videos were all about how to find a confidence interval, and not what it does. I suspect that may be why my video, Understanding Confidence Intervals, remains popular.

# Introducing statistics

## Understanding Summary Statistics 5:14 minutes

Why we need summary statistics and what each of them does. It is not about how to calculate the statistics, but what they mean. It uses the shoe example, which also appears in the PPDAC and OSEM videos.

## Understanding Graphs 6:06 minutes

I briefly explains the use and interpretation of seven different types of statistical graph. They include the pictogram, bar chart, pie chart, dot plot, stem and leaf, scatterplot and time series.

## Analysing and commenting on Graphical output using OSEM 7:13 minutes

This video teaches how to comment on graphs and other statistical output by using the acronym OSEM. It is especially useful for students in NCEA statistics classes in New Zealand, but many people everywhere can find OSEM awesome! We use the example of comparing the number of pairs of shoes men and women students say they own.

## Variation and Sampling error 6:30 minutes

Statistical methods are necessary because of the existence of variation. Sampling error is one source of variation, and is often misunderstood. This video explains sampling error, along with natural variation, explainable variation and variation due to bias. There is an accompanying video on non-sampling error.

## Sampling methods 4:54 minutes 500,000 views

This video describes five common methods of sampling in data collection – simple random, convenience, systematic, cluster and stratified. Each method has a helpful symbolic representation.

## Types of data 6:20 minutes 600,000 views

The kind of graph and analysis we can do with specific data is related to the type of data it is. In this video we explain the different levels of data, with examples. This video is particularly popular at the start of courses.

## Important Statistical concepts 5:34 minutes 50,000 views

This video does not receive the views it deserves, as it covers three really important ideas. Maybe I should split it up into three videos. The ideas are the difference between significance and usefulness, evidence and strength of effect, causation and association.

Other videos complementary to these, but not on YouTube are:

• The statistical enquiry process
• Understanding the Box Plot
• Non-sampling error

# Videos for teaching hypothesis testing

## Understanding Statistical inference 6:46 minutes 40,000 views

The most difficult concept in statistics is that of inference. This video explains what statistical inference is and gives memorable examples. It is based on research around three concepts pivotal to inference – that the sample is likely to be a good representation of the population, that there is an element of uncertainty as to how well the sample represents the population, and that the way the sample is taken matters.

## Understanding the p-value 4:43 minutes 500,000 views

This video explains how to use the p-value to draw conclusions from statistical output. It includes the story of Helen, making sure that the choconutties she sells have sufficient peanuts. It introduces the helpful phrase “p is low, null must go”.

## Inference and evidence 3:34 minutes

This is a newer video, based on a little example I used in lectures to help students see the link between evidence and inference. Of course it involves chocolate.

## Hypothesis tests 7:38 minutes 350,000 views

This entertaining video works step-by-step through a hypothesis test. Helen wishes to know whether giving away free stickers will increase her chocolate sales. This video develops the ideas from “Understanding the p-value”, giving more of the process of hypothesis testing. It is also complemented by the following video, that shows how to perform the analysis using Excel.

## Two-means t-test in Excel 3:54 minutes 50,000 views

A step-by-step lesson on how to perform an independent samples t-test for difference of two means using the Data Analysis ToolPak in Excel. This is a companion video to Hypothesis tests, p-value, two means t-test.

## Choosing which statistical test to use 9:33 minutes 500,000 views

I am particularly proud of this video, and the way it links the different tests together. It took a lot of work to come up with this. First it outlines a process for thinking about the data, the sample and the thing you are trying to find out. Then it works through seven tests with scenarios based around Helen and the Choconutties. This video is particularly popular near the end of the semester, for tying together the different tests and applications.

# Confidence Intervals

## Understanding Confidence Intervals 4:02 minutes 500,000 views

This short video gives an explanation of the concept of confidence intervals, with helpful diagrams and examples. The emphasis is on what a confidence interval is and how it is used, rather than how they are calculated or derived.

## Calculating the confidence interval for a mean using a formula 5:29 minutes 200,000 views

This video carries on from “Understanding Confidence Intervals” and introduces a formula for calculating a confidence interval for a mean. It uses graphics and animation to help understanding.

There are also videos pertinent to the New Zealand curriculum using bootstrapping and informal methods to find confidence intervals.

# Probability

## Introduction to Probability 2:54 minutes

This video explains what probability is and why we use it. It does NOT use dice, coins or balls in urns. It is the first in a series of six videos introducing basic probability with a conceptual approach. The other five videos can be accessed through subscription.

## Understanding Random Variables 5:08 minutes 90,000 views

The idea of a random variable can be surprisingly difficult. In this video we help you learn what a random variable is, and the difference between discrete and continuous random variables. It uses the example of Luke and his ice cream stand.

## Understanding the Normal Distribution 7:44 minutes

In this video we explain the characteristics of the normal distribution, and why it is so useful as a model for real-life entities.

There are also two other videos about random variables, discrete and continuous.

## Risk and Screening 7:54 minutes

This video explains about risk and screening, and shows how to calculate and express rates of false positives and false negatives. An imaginary disease, “Earpox” is used for the examples.

# Other videos

## Designing a Questionnaire 5:23 minutes 40,000 views

This was written specifically to support learning in Level 1 NCEA in the NZ school system but is relevant for anyone needing to design a questionnaire. There is a companion video on good and bad questions.

# Line-fitting and regression

## Scatterplots in Excel 5:17 minutes

The first step in doing a regression in Excel is to fit the line using a Scatter plot. This video shows how to do this, illustrated by the story of Helen and the effect of temperature on her sales of choconutties

## Regression in Excel 6:27 minutes

This video explains Regression and how to perform regression in Excel and interpret the output. The story of Helen and her choconutties continues. This follows on from Scatterplots in Excel and Understanding the p-value.

There are three videos introducing bivariate relationships in a more conceptual way.

There are also videos covering experimental design and randomisation, time series analysis and networks. In the pipeline is a video “understanding the Central Limit Theorem.”

# Supporting our endeavours

As explained in a previous post, Lessons for a budding Social Enterprise, Statistics Learning Centre is a social enterprise, with our aim to build a world of mathematicians and enable people to make intelligent use of statistics. Though we get some income from YouTube videos, it does not support the development of more videos. If you would like to help us to create further videos contact us to discuss subscriptions, sponsorship, donations and advertising possibilities. info@statsLC.com or n.petty@statsLC.com.

# The flipped classroom

Back in the mid1980s I was a trainee teacher at a high school in Rotorua. My associate teacher commented that she didn’t like to give homework much of the time as the students tended to practise things wrong, thus entrenching bad habits away from her watchful gaze. She had  a very good point! Bad habits can easily be developed when practising solving equations, trigonometry, geometry.

Recently the idea of the “flipped classroom” has gained traction, particularly enabled by near universal access to internet technology in some schools or neighbourhoods. When one “flips” the classroom, the students spend their homework time learning content – watching a video or reading notes. Then the classroom time is used for putting skills to practice, interactive activities, group work, problem-solving – all active things that are better with the teacher around. Having a teacher stand at the front of the room and lecture for a large percentage of the time is not effective teaching practice.

I ws surprised at a teaching workshop to find that many of the teachers were not even aware of the concept of “flipping”. To me this is a case for Twitter as a form of professional development. To address this gap, I am writing about the flipped classroom, especially with regard to statistics and mathematics.

There are two important aspects to flipping – what the students do when they are not in class, and what students do when they are in class.

## Work away from class

In theory, classroom “flipping” has always been possible. You could set students notes to read or sections of the textbook to study. In some schools and cultures this is successful, though it does presuppose a high level of literacy. Universities expect students to read, though my experience is that they avoid it if possible – unless they are taking Law, which of course means they can’t avoid it.

Technology has changed the landscape for flipping. With ready access to the internet it is feasible for video and other work to be set remotely for students. Sometimes teachers prepare the material themselves, and sometimes they may specify a YouTube video or similar to watch. This is not as easy as it may sound. As you can see from my critique of videos about confidence intervals, there is a lot of dross from which to extract the gold. And Khan Academy is no exception.

One big advantage of video over a live lecture, even if the video is merely a talking head, is that the student can control the pace and repeat parts that aren’t clear. My experience of lecturing to classes of several hundred students was that the experience was far from personal. I would set the pace to aim at the middle, as I’m sure most do. In later years I put all my lectures into short audio files with accompanying notes. Students could control the pace and repeat parts they didn’t understand. They could stop and think for a bit and do the exercises as I suggested, sometimes using Excel in parallel. They could quickly look through the notes to see if they even needed to listen to the audio. It was much more individualised.

Another advantage was that you can remove errors, stumbles, gaps and tighten up the experience. I’ve found a fifty minute lecture can be reduced to about half the time, in terms of the recording.

Despite this much more individual approach I was still expected to give lectures (that’s what lecturers do isn’t it?) until the Christchurch earthquakes made my mode of delivery expedient and we were able to stop physical lectures. The students could view the delivery of the material without coming to the university. They could then do exercises, also set up on the LMS, with instant feedback.

# Work in the classroom

People tend to focus on what happens away from the classroom, when talking about flipped classroom. It is equally important to think hard about what happens in class. Having the teacher and peers there to help when working through problems in mathematics is better than being at a dead-end at home, with no one to help.  But week after week of turning up to class, working on numbered exercises from the textbooks doesn’t sound like much fun.

Taking the content delivery out of the classroom frees up the teacher for all sorts of different activities. It can be a challenge for teachers to change how they think about how to use the time. There are opportunities for more active learning, based on the grounding done on-line. In a mathematics or statistics classroom there is room for creativity and imagination. Debates, group work, competitions, games, looking for errors, peer review and peer-grading are all possibilities. If anyone thinks there is no room for imagination in the teaching of mathematics, they should take a look at the excellent blog by Fawn Nyugen, Finding ways to Nguyen students over.  I wish she had taught my sons. Or me. (Nguyen is pronounced “Win”)

I am currently working with teachers on teaching statistical report-writing. This is something that benefits from peer review and discussion. Students can work separately to write up results, and then read each other’s work. This is done in English and Social Science classes, and language classes. There is much we can learn from teachers in other disciplines.

# Potential Problems

Students can also be resistant to change, and some coaching may be needed at the start of the year.

There is a big investment by teachers if they wish to create their own materials. Finding suitable materials on line can take longer than making your own. A team approach could help here, where teachers pool their resources and provide the “at home” resources and links for each other’s classes. I would be cautious not to try to do too much at once in implementing “flipped classroom”. It would probably be wise to start with just one class at a time.

Where internet access is not universal, there needs to be adaptations. It may be that the students can use school resources out of school time. Or students could take the material home on a memory stick.

# Special needs

One issue to consider is the students who have special learning needs. In one Twitter discussion it was suggested that the flipped classroom is great because the student can learn the content when they have a helper (parent!) to assist. This is an admirable theory and I might have agreed had I not been on the other side. As a mother of a son with special needs, the thought of homework was often too much for me. The daily battle of life was enough without adding further challenge. In addition my son had been full-on all day and had little capacity for homework even if I had been willing. We need to avoid assuming ideal circumstances.

# Try it!

Overall though, in appropriate circumstances, the concept of flipping has a lot going for it. It is always good to try new things.

If you never have a bad lesson or a failed new idea, you aren’t being daring enough!