What is Central Tendency – An Introduction to Mean, Median, and Mode in Statistics (5-1) - YouTube

Channel: Research By Design

[0]
Hi, I'm Dr. Todd Daniel and you're watching Statistics for the Flipped
[12]
Classroom. Last time, I showed you how to display data with pictures. Now, I'm going
[18]
to teach you how to display your data with numbers. There are some specific
[23]
numbers that we want to know about our data, and foremost we want to know where
[27]
is the center of the data. There are three ways that we can describe central
[32]
tendency: mean, median, and mode. I'm going to teach you when we will use each
[39]
measure of central tendency, and how to compute each one.
[43]
Let's get started.
[47]
We have two important questions about our data set. The first question is; what
[54]
single number best represents our data? Our second question is whether the
[60]
scores are packed together or spread out. The answer to our first question about
[66]
the most representative score is going to be some measure of central tendency.
[72]
A measure of central tendency indicates where the center or the middle of the
[76]
distribution tends to be. So you took a test, and you got 94%.
[83]
You may have done exceptionally well,
[85]
or maybe not. Perhaps you're really smart or perhaps the test was really easy.
[92]
If I just tell you the raw scores; 80, 94, 73, 88, 86, 72, then you still don't really
[98]
know that much, but if you know that you scored 94 and the class average was 80.5
[105]
then you know that you did really well.
[110]
A measure of central tendency answers the question of whether the scores are
[114]
generally high or generally low. For example, if I tell you about my friend
[120]
who is a basketball player,
[122]
what do you assume about his height? Based on what you know about basketball
[128]
players in general, you might assume that his height is above average because you
[134]
know that the height of basketball players is generally high. If I tell you
[138]
about my friend who is a horse jockey, you might assume that his height is
[143]
generally low. These estimates are based on what you already know about what the
[149]
heights of certain individuals tend to be. You are using a measure of central
[155]
tendency. Measures of central tendency offer the benefit of simplification.
[162]
The whole sample of raw data is difficult to understand. A measure of central tendency
[167]
gives us a single number that represents the entire data set. Central tendency
[173]
also has the benefit of prediction. Knowing a measure of central tendency
[178]
can help us to predict other scores
[181]
or estimate future needs and this is true even if the measure of central
[186]
tendency is not in the actual data set. If I tell you that the average family
[191]
has 2.3 children, but that our city is projected to grow by 1,000 families in
[198]
the next five years, how many new classroom seats will we need in public
[203]
schools? Although no family who moves to our town literally has 2.3 children, that
[211]
average is still very useful for predicting how many children will be
[215]
entering the school system. Determining which measure of central tendency to use
[220]
depends upon two things.
[223]
First, it depends upon the scale of measurement; nominal, ordinal, interval, or ratio.
[230]
For instance, the mean is typically used with interval or ratio level data. Second, our
[238]
choice of a measure of central tendency depends upon the shape of the
[241]
distribution. If the distribution is highly skewed or kurtotic, we may choose a
[247]
specific measure of central tendency. There are three measures of central
[252]
tendency that we will learn about; mean, median, and mode.