Human cognition influence on the OSM data classification Study
Posted by loai_84 on 27 November 2017 in English (English). Last updated on 28 November 2017.Dear participants,
Thank you for helping science and supporting voluntary geographic contents. The contents which have been exclusively collected by voluntary efforts. In these contents, public with local knowledge is the main source of information. Thus, we are (as a computer scientist) concerning with human cognition influence on data classification. To participate in the study press here
For example, regarding landuse/landcover classification, we could agree on that a parcel of land is covered by grass or water which is an abstract level of classification. However, whether this parcel is a park, a garden, a cemetery, or even a golf court might be perceived differently. Whether this water body is a pond, a lake or a reservoir likely requires finer measures, geographic knowledge, and particular expertise.
Therefore, we developed a study to understand more about how people perceive geographic contents remotely. The study aims to:
- find out human capabilities to provide landuse data.
- understand the influence of classification mechanism and schema on data quality.
- study human behaviors in providing voluntary data, and
- emphasis challenges of data classification in voluntary contents.
The study takes about 25-30 min. We do not collect any personally identifiable information. Thank you again for your time and your contribution. To participate in the study press here
Don’t hesitate to contact me for further comments and feedbacks.
Best regards, Dr. Ahmed Loai Ali loai@informatik.uni-bremen.de
Comment from wanderlustig on 3 December 2017 at 03:09
The test is very detailed. 25-30 minutes is optimistic. 60 minutes is more realistic.
Also, the scale of the Performance factor on the last two pages has the opposite magnitude to the other scales (right = low). I realise this is the NASA-TLX standard, but you should draw attention to it (e.g. by bolding “High” and “Good” under each scale) because people who started the test with the expectation of half an hour may be under intense time pressure by the time those last two pages pop up.
Finally, there’s no way to go back and correct earlier answers.
Please post an update once you publish your findings. It will be interesting to read about how people do on this test.
Comment from loai_84 on 4 December 2017 at 08:24
Thank you very much for your comments
I feel sorrow that our time estimation makes you feel optimistic. However, thank a lot for your time and your contribution to the study. Regarding the TLX performance measure, I have the same misunderstanding in the beginning until I realize why it looks like that. I will update the study to hint the user for that.
I will keep you updated with the result sure. But, If you could make us a favor by publishing the study to any channel or community. It would be great, and super contribution from your side.
Thanks again, Ahmed loai ali
Comment from wanderlustig on 4 December 2017 at 08:37
No problem.
You might want to post a link to this page on the OpenStreetMap subreddit.