Post on 21-Jan-2016
Synthesizing Data: from noise to music
Fiona Jeffries, MABrandon Regional Health Authority
Data Leading to Action: From Chaos to ClarityJanuary 19, 2009
Where’s my circuit board?
No magic formula
Ask more questions! Of your data Of other sources Of the people described in the data
Start with your data Look at your numbers What stands out? Are they what you expected? Why might they be the way they are?
Start with your data - example
Body Weight
20% 17%
63%
22% 10%
67%
0%
20%
40%
60%
80%
100%
Underw eight Healthy w eight Overw eight
% o
f res
pond
ents
M F
Perception of Body Weight
57%
23%18%
54%
11%
32%
0%
20%
40%
60%
80%
100%
Underw eight Healthy w eight Overw eight
% o
f res
pond
ents
M F
Question your data
Multivariate analysis? Compare related data within your
data set
What are your assumptions? Do the data seem to confirm them or
suggest something else? What other questions occur to you?
Question your data - example
Multivariate analysis Does overweight correlate with:
Inactivity Low Fruit and Vegetable consumption Feeling safe at school?
What links can we make with other indicators?
Correlation ≠ Causation
Look to the sources What other sources speak to your
data? Population surveys (e.g. CCHS, Census)
Likely more general than your data
Published research (academic journals) Likely more specific than your data
Differences or similarities? Do they answer your questions or suggest
more questions?
Look to the sources - example CCHS 2004
Canadian household population age 12-17 29% overweight
Published Research Abnormal perception of body weight is not
solely observed in pubertal girls: incorrect body image in children and its relationship to body weight, H. Ohtahara, et al. Acta Psychiatrica Scandinavica Volume 87 Issue 3
Ask the subjects
Qualitative data – what people say Ask the people you studied what they
think
Focus groups World Café discussions Photovoice Participatory video Social networking groups
The circuit board Your brain – think about what you find Your team – discuss with others who
may know things you don’t know Explore possible meanings and ways
the story can fit together
Loop! Repeat the steps as needed – this is not
a linear process!
Play the tune
The story begins to emerge Show your data Describe how the other sources add
to the story Don’t be afraid to point out questions
you can’t answer!