Analyzedata Frequencytables

You might also like

Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 6

Frequency Tables

You have your results, now we need to turn this data into a more useable format to
determine unit preferences.

The first step is to produce a frequency table. A frequency table is a tabulation of the
responses organized in such a way that you can determine the actual total responses
for your questions. So for our example one way we can set up the frequency table is as
follows:

Sectio Coffee Tea Total


n
HQ
A Div
B Div
Ops
Total

As you review your raw data (the circle responses from your colleagues), add a “tick
mark” for each response. For example for the first response (A Div, Tea) you would
place a “tick mark” in the box that is beside A Div and under Tea. Repeat this until you
have captured all of your “tick marks”.

Section Coffee Tea Total


HQ I II
A Div II IIII
B Div I IIII
Ops III I
Total

Then you can convert the ticks into numbers as follows.

Section Coffee Tea Total


HQ I1 II2
2
A Div II IIII4
B Div I1 IIII5
Ops III3 I1
Total

Once you have your numbers you can then generate all of your totals, your table should
end up looking like this one.
Section Coffee Tea Total
HQ 1 2 3
A Div 2 4 6
B Div 1 5 6
Ops 3 1 4
Total 7 12 19

We can convert this tabular data into a chart called a “histogram” as follows.

4
Coffee
3
Tea
2

0
HQ A Div B Div Ops

Now we have created actual numbers we can work with to determine unit preferences.

The first thing we need to determine is whether or not our response rate is giving us
useable information. For example, if only 5 out of 20 members of the unit participated in
the survey, is the information accurate? We can answer this by referring to the table of
Representative Sample Size found in CFITES Vol 11-1 Part 3, pg 8.

What would be the minimum number of respondents necessary to attain a statistically


representative sample for our unit of 20 members?

Answer -> 19

So based on this then, did we get enough responses for us to be confident that our data
is good?

Answer -> Yes

So, we can be confident that our data can be used to represent the unit, now what?
Well, we are still trying to determine whether the members of the unit prefer purchasing
canteen coffee or tea. We can determine this by looking at our frequency table.
What percentage of all respondents prefer purchasing canteen tea? (hint: take the total
responses for tea and divide it by the total number of responses, then multiply your
answer by 100)

You should get approximately 63%

12/19 = 0.631,
0.631 x 100 = 63%.

What other information can we get from the frequency table?

The section(s) that prefers canteen coffee the most? (Ops. 3/4 = 0.75, 0.75 x 100 =
75%) The section(s) that prefers canteen tea the most? (B Div, 5/6 = 0.83, 0.83 x
100 = 83%)

Central Tendency

If you have a data set like this where the answers provided are based on selecting an
item, rather than selecting from a scale of responses, then the approach taken above is
recommended. However, if you have asked questions of your respondents that require
them to provide a response based on a scale then you will need to use the concept of
“central tendency” to better understand your data. So the one thing we haven’t taken
into account yet is not just preference, but how much canteen coffee or tea people
actually drink. This will tell us how much inventory we need. In order to understand this
we need a different type of question based on a scale.

For example, in our unit we could have asked the question about preference as follows:

1. On the following scale how many coffees from the unit canteen do you purchase per
day?

a. 1
b. 2
c. 3
d. 4
e. 5
f. 6 or more

2. On the following scale how many teas from the unit canteen do you purchase per
day?

a. 1
b. 2
c. 3
d. 4
e. 5
f. 6 or more

So let’s assume that you did include these questions in your survey, here are the
responses.

Quantity Coffee Tea


1 1 3
2 1 3
3 1 2
4 1 2
5 0 1
6 3 1
Total 7 12

So is this all the data we need in order to know how much coffee and tea is consumed
by our section? Well not quite… we need to determine what the “central tendency” of
these responses is so that we can work out how much inventory we will require.

The central tendency of a set of numbers is an estimate of the “center” of the values.
There are three ways by which we estimate the center of a range of values, they are:

1. Mean – the average of all of the numbers.

2. Median – the absolute middle value in a range of values when they are
lined-up from smallest to largest. When you have an even range of values,
then you take the average of the two middle values.

3. Mode – is the most frequent value that appears in a range of values

So the data set we have to use for the coffee drinkers would be as follows:

1,2,3,4,6,6,6

MEAN

Therefore, our MEAN number of cups purchased would be:


1+2+3+4+6+6+6 = 28, we know that we only have 7 coffee drinkers, therefore our
average would be: 28/7 = 4 cups of coffee a day per coffee drinker. So, our
MEAN for coffee drinkers is 4 cups per person per day.

Okay, so what would be the MEAN number of cups purchased by the tea drinkers?
(hint: data set is: 1,1,1,2,2,2,3,3,4,4,5,6)
(Answer = 2.8 cups per person a day)

MEDIAN

So for the coffee data set the MEDIAN would be as follows:


The center of this range is as indicated (odd number of values returns a single value)
1,2,3,4,6,6,6 (Since we have 7 values the middle one is 4 cups per person a day)

So what would the MEDIAN be for the tea data set?


(Hint: Since it is an even number set, we will have to take the average of the two middle
values…)
(Answer = 2.5 cups per person a day)

MODE

The MODE for the coffee data set would be the most frequently occurring value in this
case it would be 1,2,3,4,6,6,6, therefore MODE = 6.

So what would the MODE be for the tea data set?


(hint: you can have more than one)
(Answer, MODE = 1 and 2)

Alright, so now that we have calculated all this, what does it mean?

Coffee Drinkers Tea Drinkers


MEAN = 4 MEAN = 2.8
MEDIAN = 4 MEDIAN = 2.5
MODE = 6 MODE = 1 and 2

Since these measures are estimates of the “center” of the range of values, in the case
of the coffee drinkers, what do these estimates tell us? How much coffee purchased per
person per day should be held in our inventory? Which number would you go by
(MEAN, MEDIAN, MODE)?

There is actually no wrong answer here as long as you don’t pick something less than 4
or more than 6. The objective is to have enough coffee in your inventory to last a week
at a time. You don’t want to keep too much, which is what you risk if you use the MODE.
Conversely you don’t want to run-out which might happen if you go with the MEAN or
MEDIAN. Because at this point we are subjectively viewing this data you could even
decide to go with an inventory level that would give 5 cups per person per day to
account for the 6 or more cups a day “coffee fiends” that the MODE represents.

Given the process we just went through for determining the canteen coffee inventory,
what would you determine the canteen tea purchase rate per person per day to be?

(Answer: If you revisit the data set, you will notice that the MODE is not realistically
representing the group of tea drinkers, because even though there are more people that
drink either 1 or 2 cups a day, 50% of the group are consuming 3 or more cups a day.
In this case, the MEAN = 2.8 cups per person per day, would provide a more suitable
estimate of the consumption habits of Unit tea drinkers.)

So now that we’ve looked at the QUANTITATIVE approach (numbers driven) we need
to now look at the QUALITATIVE approach (open-ended questions).

You might also like