Transcript 1.4 Believing Statistics??

Statistical Reasoning
for everyday life
Intro to Probability and
Statistics
Mr. Spering – Room 113
1.4 Believing Statistics??

Check IN????
 Have
you ever heard an unbelievable statistic
or claim? How did you react?
 Research
Shows????? Show me the
research? When should you believe a
statistical study?
1.4 Believing Statistics??




Most researchers conduct studies with honesty
and integrity
However, bias can arise in many ways. Thus we
examine statistics carefully, looking for anything
that might cause question. “BE METICULOUS!”
No definitive way to ask “Should I believe a
statistical study?”
Nevertheless, the following guidelines can be
helpful.
1.4 Believing Statistics??

Eight Guidelines for Evaluating a Study
1.
2.
3.
4.
5.
Identify the goal, population, and type of study
Consider the source, are they biased
Question the sampling method, “Is it
representative?”
Look for problems in defining or measuring
variables of interest
Watch for confounding factors that can invalidate
conclusions
TO BE CONTINUED….
1.4 Believing Statistics??

Eight Guidelines for Evaluating a Study
6.
7.
8.
Consider any survey, looking for anything that might
produce inaccurate or dishonest results
Check that results are fairly represented in graphs
and concluding statements, both researchers and
media often create misleading graphics or jump to
conclusions not supported by results.
Finally, consider the conclusions. Do the
conclusions make sense? Do the results have any
particular significance? Did the study achieve its
goals?
Use your own judgments
to extend these guidelines!
1.4 Believing Statistics??

Bizarre Factoid:
The only guy without a beard in the band ZZTOP is….

Statistic:
3% of pet owners give
Valentine's gifts to their pets
on Valentine's Day.
 ???

Frank
Beard
1.4 Believing Statistics??

Consider the source: Help prevent
problem studies by using Peer Review.
review – Process in which several
experts in a field evaluate a research report
before the report is published.
 Peer
1.4 Believing Statistics??


Examine the Sampling Method: Poor sampling
methods almost guarantee bias. (GALLUP)
Check for these confounding factors –
bias (selection effect) – occurs when
sample chosen in biased way
 Participation bias – occurs when subjects in study are
voluntary
 Voluntary response survey – remember self-selected
survey and convenience sampling
 Selection
1.4 Believing Statistics??



Variable – item or quantity that can vary or take
on different values
Variables of interest – items or quantities that the
study seeks to measure
Problems in defining or measuring variables:
Results of statistical studies are harder to
interpret and seem viable if the variables under
study are difficult to define or measure
1.4 Believing Statistics??
Watch out for confounding variables:
Often, other variables that are not part of the study
can make it difficult to interpret results.
Confounding factors → confounding variables
Confounding variables can affect a study’s results if
they have not been accounted for, they are difficult
to discover, and sometimes they are never
discovered.

1.4 Believing Statistics??
Consider the wording of a survey:
Unfortunately most people do not have a problem
with lying → avoid sensitive subjects
Availability error – people are more likely to
choose the item that comes first in a survey
because of a tendency to make judgments on
what is most available in the mind.

1.4 Believing Statistics??

Stand back and consider the
conclusions:
 Did
the study achieve its goals?
 Do the conclusions make sense?
 Can you rule out alternative explanations?
 If conclusions make sense, do they have
significance? (i.e. losing half a pound, killing
5 bacteria, saving 2 minutes)
Extraordinary claims require extraordinary evidence!
-Carl Sagan
1.4 Believing Statistics??
1.4 Believing Statistics??

Why is a self-selected survey always
prone to bias?
People who feel strongly about an issue are more
likely to respond, therefore the opinions of the
respondents is unlikely to represent the opinions of a
larger population.

What are confounding variables?
Confounding factors
1.4 Believing Statistics??


HOMEWORK # 4:
pg 42 # 6-30 even
BE INQUSITIVE!