Why research (and researchers)
won’t always tell you the truth
Why researchers sometimes go wrong
•

They make mistakes

•

They talk to the wrong people (e.g. they can’t talk
to people without landlines on CATI surveys)

•

They confuse cause with effect (A and B regularly
happen together – therefore A causes B)

•

They are sure of the conclusion before the research
– & design the research (or fudge the data) to prove it

•

They are frightened of the wrong conclusions

•

They have reasons to lie (e.g. ties to funding
organisations)
Questionable research practices are common

John et al. (2011) Prevalence of Questionable Research Practices. Psychological Science 23(5) 524–532

3
Why don’t people always tell the truth?
•

They want to be liked

•

They are boasting (especially on social media!)

•

They don’t want to feel stupid

•

They hope to win something

•

They get bored

•

They can’t be bothered to think

•

Or perhaps they do tell the truth but they are not a
representative sample
•

Social media participants and survey respondents
are often “people with something to say”
How not to tell the truth in surveys

Sir Humphrey Appleby knew all about leading questions!
http://www.youtube.com/watch?v=G0ZZJXw4MTA
Selecting the data
•

What is the “best” set of data to fit my thesis

•

Are there “outliers” I can ignore – or include
Selecting the axes

7
Data visualisation – the wrong chart

Pie charts can easily cause
problems

And infographics can be very difficult to get right
(area of B on the left is 9x that of B on the right)
Selecting what you report
What is the average income in my street?

1 teacher
earning
One celeb earning
14 managers
£35,000
£4,464,000
making £50,000
20 other people only
Five professionals
making £10,000
making £150,000

mean = £150,000; median = £35,000; mode = £10,000
Conclusions

Handle data with care
Business is an art, not a science!
jeremy@mosoco.co.uk

10

Research and truth

  • 1.
    Why research (andresearchers) won’t always tell you the truth
  • 2.
    Why researchers sometimesgo wrong • They make mistakes • They talk to the wrong people (e.g. they can’t talk to people without landlines on CATI surveys) • They confuse cause with effect (A and B regularly happen together – therefore A causes B) • They are sure of the conclusion before the research – & design the research (or fudge the data) to prove it • They are frightened of the wrong conclusions • They have reasons to lie (e.g. ties to funding organisations)
  • 3.
    Questionable research practicesare common John et al. (2011) Prevalence of Questionable Research Practices. Psychological Science 23(5) 524–532 3
  • 4.
    Why don’t peoplealways tell the truth? • They want to be liked • They are boasting (especially on social media!) • They don’t want to feel stupid • They hope to win something • They get bored • They can’t be bothered to think • Or perhaps they do tell the truth but they are not a representative sample • Social media participants and survey respondents are often “people with something to say”
  • 5.
    How not totell the truth in surveys Sir Humphrey Appleby knew all about leading questions! http://www.youtube.com/watch?v=G0ZZJXw4MTA
  • 6.
    Selecting the data • Whatis the “best” set of data to fit my thesis • Are there “outliers” I can ignore – or include
  • 7.
  • 8.
    Data visualisation –the wrong chart Pie charts can easily cause problems And infographics can be very difficult to get right (area of B on the left is 9x that of B on the right)
  • 9.
    Selecting what youreport What is the average income in my street? 1 teacher earning One celeb earning 14 managers £35,000 £4,464,000 making £50,000 20 other people only Five professionals making £10,000 making £150,000 mean = £150,000; median = £35,000; mode = £10,000
  • 10.
    Conclusions Handle data withcare Business is an art, not a science! jeremy@mosoco.co.uk 10

Editor's Notes

  • #2 elderly Taiwanese people who shop every day are 27% less likely to die over 10 years than those who shop once a week –cause or effect?