Investigation on the effects of data manipulation

1
Investigation on the effects of data manipulation

• Part 1 Research on Gun Deaths

Part 2 Graphing and data manipulation
Researchers Matt Johns, Will Parkinson and
James Pye for Mr. Lieff, Carleton Place High
School
2
Part 1 Gun death Research

• Initially we attempted to obtain data on the
  relations between gun deaths and various trends
  in Canada.
• We will show a portion of the data we collected.
• The research will investigate trends and
  influences that certain government bills have had
  on gun deaths in Canada.

3
(No Transcript)
4
(No Transcript)
5
Bill C-150

• The passage of Bill C-150 was May 14th, 1969
• The Bill deals with abortion rights.
• The moral and political content was downplayed.

• It legalized contraception and therapeutic
  abortion, but it assigned doctors control over
  women's reproductive decisions.

6
Bill C-51

• Enforced since 1977
• This Act requires acquisition certification for
  all firearms, restricts the availability of some
  types of firearms to certain types of
  individuals.
• It establishes procedures for handling and
  storing firearms, requires permits for those
  selling firearms, and increases the sentences for
  firearm offences.

7
Bill C-17

• The overarching goal of both the previous bill
  and this bill created in 1991 named Firearms
  Control Initiative was meant to enhance public
  safety from firearms accidents and suicides as
  well as reducing criminal firearms misuse thereby
  making Canada a safer society.

8
Bill C-68

• The bill imposes a mandatory minimum sentence of
  four years in prison, in addition to a lifetime
  prohibition against possession of a firearm upon
  conviction of any of ten specific violent
  offences with a firearm.
• Proclaimed into law on January 1, 1996.

• The offences affected are manslaughter,
  attempted murder, sexual assault with a weapon,
  robbery and criminal negligence causing death,
  causing bodily harm with intent, aggravated
  sexual assault, kidnapping, hostage taking,
  extortion.

9
Part 2 An Investigation on Data Manipulation.

• We will show how anybody can manipulate data to
  display any trend they wish.
• We will use different graphing techniques and
  mathematical functions to display the data.

10
Changing the Scale
11
(No Transcript)
12

• By Changing the scale on a graph we are able to
  change the appearance of a relation.
• As seen the previous graph the was a line with a
  R2 of 1.0 and slope of 1000.
• In the next graph observe how the slope and
  equation remain the same although the graph looks
  much different.
• By changing the scale on the Y axis you can
  change the appearance of the slope of a graph.

13
(No Transcript)
14
Small Sample Size
15
(No Transcript)
16
Too Few Data Points

• Here it shows that a large immigration population
  leads to a large amount of gun deaths but once
  the immigrant population reaches 1.4 million the
  gun deaths begin to decrease.
• There is obvious problem with the data in the way
  in which it is presented because this relation is
  improbable.
• As you can see there are only 4 data points on
  this graph that the curve is fitting to which
  makes the curve easily fit the data.
• This explains why this curve only has as R2 of
  0.96

17
The Log Function
18
(No Transcript)
19
Y 1/Log(x)

• By using the logarithmic function on x we can
  change the appearance of the graph.
• Although the equation changes the data that is
  used remains the same, it is just placed into the
  equation y 1/log(x)
• This is used to show how the U.S. Government debt
  does not appear to be increasing at a exponential
  rate like it actually is.

20
(No Transcript)
21
Logarithmic Scale
22
(No Transcript)
23
(No Transcript)
24

• As you have noticed the scale on the x-axis of
  the graph has a very weird structure.
• This is called a logarithmic scale or a semi-log
  graph.
• 1 to 10 takes up the same amount of space as 10
  to 100. Similarly 10 to 100 takes up the same
  space as 100 to 1000.
• Doing this changes a linear equation into a
  equation with a curve.
• In this case it makes the data look like it is
  decreasing at a faster rate than it really is.

25
Coincidences
26
Population vs. Bacon Prices
27

• You can use a coincidence to find a relation that
  shouldnt exist.
• Using this people can draw false conclusions on
  the data.

28
Conclusion

• We have shown that data can be manipulated to
  show any relation needed to be expressed.
• All that needs to be done is footnote what has
  actually happened
• By doing this most people wont pick up on the
  footnote and the people who made the graph cannot
  get in trouble for falsified data
• You should always look for stars that lead to
  footnotes to truly understand the data.