Newsflash! Errors in the news!

NEWSFLASH!
IS THE NEWS ACCURATE?
Laura Jennings and Timina Liu
Year 11
Burgmann Anglican School
Aim: To determine the probability of errors
on each page of a newspaper using the
Poisson Distribution.
Hypothesis: It is expected that there will
be less than 5 errors on average on a page
of a newspaper
Materials:
ο‚’ 3 newspapers of the same publisher
ο‚’ Pen
ο‚’ Excel
Method:
1. All the materials were gathered.
2. The first 10 pages with articles in a
newspaper were carefully checked for
spelling, grammatical and formatting
errors. All errors were circled with a pen.
3. The number of errors on each of the pages
were recorded, with their relative
frequency calculated, plotted and fitted by
a Poisson Distribution.
4. Steps 2 and 3 were repeated for the other 2
newspapers.
Results:
All count data were used to find the average
number of errors to be substituted into the
Poisson Distribution formula to calculate the
probabilities:
𝑒 βˆ’πœ† πœ†π‘₯
𝑃 π‘₯ =
π‘₯!
where πœ† = average no. of errors and
π‘₯ = 0,1, 2, 3, … (0! = 1)
The probabilities were graphed.
The orange dots represent the Poisson probability
curve. The blue dots represent the relative counts
for the raw data collected.
Probability of errors on a page of
Newspaper 1
0.35
0.3
0.25
Probability
The graph on the right
(with average = 4.3) shows
that Newspaper 1 has the
highest probability of 4
errors on a page.
There is a small probability
of having 0 or more than 10
errors.
0.2
0.15
0.1
0.05
0
0
2
0.45
0.4
Probability
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
2
4
6
8
Number of errors on a page
10
6
8
10
12
Number of errors on a page
Probability of errors on a page of
Newspaper 2
0
4
12
The graph on the left (with
average =5.2) shows that
the probability of having 5
errors on a page is the
highest, with a small
chance of having 0 or more
than 10 errors.
Probability of errors on a page of
Newspaper 3
0.45
0.4
0.35
Probability
The graph on the right
(with average = 0.9) shows
that the probability of
having no errors on a page
is the highest, with a small
chance of having 5 or more
errors.
0.3
0.25
0.2
0.15
0.1
0.05
0
0
2
Probability
0.2
0.15
0.1
0.05
0
4
6
8
Number of errors on a page
10
8
10
12
The graph on the left (with
average = 3.5) shows that
for all 3 newspaper, the
probability of having 3
errors on a page is the
highest, with having 10 or
more errors being unlikely.
0.25
2
6
Number of errors on a page
Probability of errors on a page
of all newspapers
0
4
12
Discussion
ο‚’ The general trend of the results has fitted
well. The probabilities provide useful
information on the number of errors in the
newspapers.
ο‚’ There are a small number of inconsistencies
between the relative frequencies and the
probability curve. However, they can be fixed.
ο‚’ More pages for counts for the 3 newspapers
can be used in the analysis.
ο‚’ Further comparison could have been made
between the different sections in the
newspapers.
Conclusion
ο‚’ The results support our hypothesis.
ο‚’ For the majority of the results, the
curve peaks at 3 to 5 errors per page.
ο‚’ It is clear that it is rare to have no
errors or to have more than 10 errors
per page.
ο‚’ Newspaper 3 has the smallest average
of errors and is therefore better than
the other two newspapers for the
period of data collection.