More creative ways to present statistical results / data
y-axis
x-axis
or “the worst graphs ever” !?
The next examples are taken from a web-pagethat shares educational material for teachers(the graphs were actually published in newspapers and magazines)
http://dpcdsb-gains.wikispaces.com/file/view/Worst+Graphs+Ever.pdf/126543183/Worst%20Graphs%20Ever.pdf (retrieved March 2014)
More creative ways to present statistical results / data
y-axis
x-axis
Note: When talkingabout regression
We say“y is regressed on x”
y-axis
x-axis
1955 1965
Δx=10yr
Δ
Δx=2yr
1973 1975
y-axis
x-axis
1955 1965
Δx=10yr
Δ
Δx=2yr
1973 1975
$58,000
$50,000
$16,000
$29,000
Δy=$8,000
Δy=$13,000
Δy/Δx=$8,000/2yr
Δy/Δx=$13,000/10yr
1940 1960 1980
$20,000
$40,000
$60,000
Distance (y-axis)
Time (x-axis)
Δy=0.5mile
Δy=4miles
Distortion factor ( ‘Lie-factor’)
And the objective presentation of the data
Some more creative ways to hide or distort the statistical results…
El Niño region SST departures (anomalies) (oC)
measured in different regions of the tropical
Pacific
Climate Variability:El Niño - Southern Oscillation
(SST: Sea Surface Temperature)
El Niño Region SST Departures (oC) Recent Evolution
Climate Variability:El Niño - Southern Oscillation
Image source: http://www.ncdc.noaa.gov/paleo/pubs/cobb2003/cobb2003.html
Fossil corals
Climate Variability:El Niño - Southern
Oscillation
Cobb, K.M., C.D. Charles, H. Cheng & R.L. Edwards, 2003,El Niño-Southern Oscillation and tropical Pacific climate during the last millennium. Nature, Vol. 424, No. 6946, pp. 271 - 276 (17 July 2003).
Red: Observed SST anomaliesBlack : Coral reconstructions(oxygen isotopes)
Climate Variability:El Niño - Southern Oscillation
http://www.ncdc.noaa.gov/paleo/pubs/cobb2003/cobb2003.html
Cobb, K.M., C.D. Charles, H. Cheng & R.L. Edwards, 2003,El Niño-Southern Oscillation and tropical Pacific climate during the last millennium. Nature, Vol. 424, No. 6946, pp. 271 - 276 (17 July 2003).
Reconstructed Climate Variability:A.D. 1320-1480
http://www.ncdc.noaa.gov/paleo/pubs/cobb2003/cobb2003.html
Effect of ENSO on Global Rainfall
http://precip.gsfc.nasa.gov/rain_pages/el_nino_vsn2.html
From Prof. Aiguo Dai’s paper in Geophysical Research Letters (2000)
Global Teleconnection Pattern
Effect of ENSO on Global Rainfall
http://precip.gsfc.nasa.gov/rain_pages/el_nino_vsn2.html
U.S. Temperature and Precipitation Departures During the Last 30 and 90 Days
30-day (ending 22 Mar 2014) temperature departures (degree C)
90-day (ending 22 Mar 2014) % of average precipitation
90-day (ending 22 Mar 2014) temperature departures (degree C)
Last 30 Days
Last 90 Days
30-day (ending 22 Mar 2014) % of average precipitation
R-scripts and data update• We will work in the next weeks with ENSO and local climate data.We will explore if we find correlations between rainfall and temperatures in
the state of New York and ENSO.Please update the following files in your local scripts-directory (If you have not done so already in class (April 27th, 2014)):
myfunctions.Rclimatology.Rplot_climatology.Rclass12.Rclass15.R
http://www.atmos.albany.edu/facstaff/timm/ATM315spring14/R/
R-scripts and data updatePlease update the following file in your local data-directory (If you have not done so already in class (April 27th, 2014)):
create a local subdirectory named ‘NY’ (for New York State)then download some of the USW station data files
NOTE: ghcnd-stations-NY.csv you open in R-studio (or text edito)To see a list of stations with geographic locations and the name of th station.
http://www.atmos.albany.edu/facstaff/timm/ATM315spring14/R/data/NY/
Processing new station data
1) Calculate the 1981-2010 climatology with climatology.R ( input is e.g. USW00094789_tavg_mon_mean.asc)
This creates two output files (a) the monthly mean climatology)( e.g. USW00094789_tavg_mon_mean_climc_1981-2010.csv)
(b) the monthly mean anomalies( e.g. USW00094789_tavg_mon_mean_ano.asc)
Processing new station data
Processing new station data2) Use plot_climatology.Rto see the climatological cycle
Processing new station data3) Use class15.RTo work the newly createdanomaly data files to comparethe time evolution and studythe correlation between twostations.
Processing new station data3) Use class15.RTo work the newly createdanomaly data files to comparethe time evolution and studythe correlation between twostations.
Note: If there are gapsin the data, the programdoes not do the calculation(this will be fixed …)