OK, so the scans are here and the data is here
If I get time I might try and do some graphs \ stats and stuff, unless anyone else wants to have a go?
cheers
steve
Structured feedback - sheets and data
- steve crawshaw
- Posts: 856
- Joined: Thu Apr 22, 2010 8:49 pm
- Contact:
Structured feedback - sheets and data
I like to keep a bottle of stimulant handy in case I see a snake, which I also keep handy.
- I_used_to_brew
- Posts: 2356
- Joined: Sat Nov 20, 2010 1:06 pm
Re: Structured feedback - sheets and data
Steve, I get the spreadsheet whichever link I choose :(
- steve crawshaw
- Posts: 856
- Joined: Thu Apr 22, 2010 8:49 pm
- Contact:
Re: Structured feedback - sheets and data
I like to keep a bottle of stimulant handy in case I see a snake, which I also keep handy.
- steve crawshaw
- Posts: 856
- Joined: Thu Apr 22, 2010 8:49 pm
- Contact:
Re: Structured feedback - sheets and data
Tinkering around with the data I have produced a chart showing overall scores (as a %) awarded by judges to each brewer's beer. This may be instructive for calibrating your approach to judging.
I will see if I can make some more charts..
cheers
steve
I will see if I can make some more charts..
cheers
steve
You do not have the required permissions to view the files attached to this post.
I like to keep a bottle of stimulant handy in case I see a snake, which I also keep handy.
- steve crawshaw
- Posts: 856
- Joined: Thu Apr 22, 2010 8:49 pm
- Contact:
Re: Structured feedback - sheets and data
A nice chart showing all the useful data on one plot. If we reduce to 5 beers I will be able to show category too. This illustrates the variation in scoring on each attribute. I have applied jitter to the points so that they don't overplot, but hopefully you can see which points belong to which attributes.
Some of the attributes (e.g. aroma and mouthfeel) will always be more grouped than others, because they are scored from a low number (3 or 5), but others, (aroma, flavour) will tend to vary more as they are marked from 12 or 20. It's interesting to note in what beers the scores for these potentially variable attributes are grouped, i.e. where is there consensus on these parameters. Paul's beer has tight grouping on most parameters, indicating perhaps that we easily approach consensus on a well - known style (8B).
For those interested in such things this was produced in the open source stats package R, using the ggplot2 library.
cheers
steve
Some of the attributes (e.g. aroma and mouthfeel) will always be more grouped than others, because they are scored from a low number (3 or 5), but others, (aroma, flavour) will tend to vary more as they are marked from 12 or 20. It's interesting to note in what beers the scores for these potentially variable attributes are grouped, i.e. where is there consensus on these parameters. Paul's beer has tight grouping on most parameters, indicating perhaps that we easily approach consensus on a well - known style (8B).
For those interested in such things this was produced in the open source stats package R, using the ggplot2 library.
cheers
steve
You do not have the required permissions to view the files attached to this post.
I like to keep a bottle of stimulant handy in case I see a snake, which I also keep handy.
Re: Structured feedback - sheets and data
Hmm, the jitter didn't come through for me so I'll score you 25/40. I did get a strong blank feeling as I stared at the plot, not knowing what I was looking at, so I'll score you 35/40 for that ... then the maths kicked in and it started to make sense.steve crawshaw wrote:I have applied jitter to the points
For example, we can see at a glance the extreme judges for any particular style. When that style next comes up whoever they may be could sit together to discuss their opinions. They may or may not find common ground.
When a man is tired of beer, he is tired of life; for there is in beer all that life can afford
Re: Structured feedback - sheets and data
Holy moly that's graphical geekiness!
The key point for me, that someone has already raised, is to have discussions after each beer.
I also thought that the number we did was too many and it is hard to flit between different styles (hence competitions use flights of specific categories). Perhaps this could be agreed before the meeting to avoid the "aah go on then" additions.
The scoring was extremely varied for some beers. I would question as to whether some are judged on personal preference (easy to do after a few drinks) rather than against the guidelines. Admittedly the latter is still subjective in approach.
The key point for me, that someone has already raised, is to have discussions after each beer.
I also thought that the number we did was too many and it is hard to flit between different styles (hence competitions use flights of specific categories). Perhaps this could be agreed before the meeting to avoid the "aah go on then" additions.
The scoring was extremely varied for some beers. I would question as to whether some are judged on personal preference (easy to do after a few drinks) rather than against the guidelines. Admittedly the latter is still subjective in approach.
James
BJCP Certified Judge
BJCP Certified Judge