KNOW THY DATA:
the basic building block
all things flow from here.
numbers simply describe the world
let them.
making charts is not so hard, finding what to make charts of can be a great challenge.
what i do-
spreadsheets are how machines read- but terrible for us.
we (PEOPLE) are pattern recognition machines- show the patterns
a small piece of advice:
GARBARGE IN, GARBAGE OUT
a small piece of advice:
GARBARGE IN, GARBAGE OUT
formats- xls, csv, tsv, sql, shp,
the dreaded PDF
where to find-
government websites
specific websties for your topic
FIOA requests
ASK YOUR SOURCES
roll your own -
http://tmsnrt.rs/2gKl5Ro
https://podio.com/site/creative-routines
vs
data is plural - good resource
explore! start a folder, bookmarks, notes, etc
follow the trails- look at bottom for sources, everyone has data, assume they do until they tell you they don't
analysis-
for larger datasets, look for databases / ask for data dictionary- KNOW THY DATA
statistics = sexy.
don't kid yourself.
a few thoughts on interpreatation:
correlation is correlation.
causation is causation.
causation is usually (linear) correlation.
but C'MON NOW
correlation is not causation!
don't get crazy
tools:
excel,
googlesheets,
sql,
sublime text,
GIS
online what have you
google is everything.
watch for:
naming conventions, date formatting, number formatting, column names
states, business names, county names, FIPS codes
cleaning issues
make things uniform to save yourself pain.
a lot of this comes with massive amounts of repitition, knowing what to look for, the warning signs.
once you have your data ready, now comes the hard part-
what does it say? how to best present?
familiarize yourself with the range of charts-