Timeline for How can I avoid being "the negative one" when giving feedback on statistics?
Current License: CC BY-SA 3.0
28 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Aug 2, 2017 at 15:17 | answer | added | Andrew B | timeline score: 1 | |
| Jan 21, 2017 at 20:07 | history | edited | Stephan Kolassa |
edited tags
|
|
| Jan 20, 2017 at 23:08 | comment | added | emory | I was just at a conference where the swag bag included a soft foam ball. We were instructed that whenever we felt a presenter was bullshitting we should throw the ball at him/her. Can you throw a ball at your colleague? | |
| Jan 20, 2017 at 1:32 | comment | added | Wildcard | @alephzero, while keeping the same basic idea, you could be less self-deprecating with e.g. "I don't see how the data supports that conclusion," or a little more firmly with "I don't see that the data supports that conclusion." | |
| Jan 19, 2017 at 22:22 | answer | added | einpoklum | timeline score: 1 | |
| Jan 19, 2017 at 14:43 | answer | added | MonkeyZeus | timeline score: 2 | |
| Jan 19, 2017 at 10:07 | vote | accept | user2390246 | ||
| Jan 18, 2017 at 17:44 | answer | added | sessej | timeline score: 23 | |
| Jan 18, 2017 at 15:02 | comment | added | Crowley | @Trilarion If it is said during, say, 2hour discussion on that topic and trying to understand it it is completely different from stating that question alone. | |
| Jan 18, 2017 at 14:59 | answer | added | Crowley | timeline score: 2 | |
| Jan 18, 2017 at 14:55 | comment | added | NoDataDumpNoContribution | @Crowley I agree that asking is the better tactic but for me the difference between the two approaches is rather small. Stating your opinion using seems/might/could ... is not the same as an outright rejection and almost the same like asking. In the course of the conversation you anyway might be forced to come up with your opinion in order to proceed. | |
| Jan 18, 2017 at 14:34 | comment | added | Crowley | @Trilarion Some people devoted many hours to form that conclusions and they are rejected without any question how (and why) they did it? Question "How did you conclude the [claim] from the [data]?" is much better - You show our interest in understanding, not in denying, and it allows you to figure why the [claim] is not justified by the [data]. | |
| Jan 18, 2017 at 14:24 | comment | added | NoDataDumpNoContribution | What's wrong with "I don't think those conclusions are really justified by the data currently."? | |
| Jan 18, 2017 at 11:36 | comment | added | Buochserhorn | In support of @alephzero's comment, in pedagogy / communication courses I came across the term I-message, which I really like. It's all about: "I don't understand why..." vs. "You must be wrong...". See here for further information: en.wikipedia.org/wiki/I-message | |
| Jan 17, 2017 at 14:51 | comment | added | ChrisLively | As a student in the field you ask the professor to explain in further detail their teaching so that you gain understanding. As a professor, you ask the student to explain their answers so that they gain understanding... | |
| Jan 17, 2017 at 14:05 | comment | added | gerrit | @alephzero When my supervisor says I don't understand why you do X, I tend to interpret it as X is wrong, but there is probably a degree of imposter effect in that interpretation. | |
| Jan 17, 2017 at 6:24 | comment | added | alephzero | The "magic words" for this are not the "please" and "thank you" you learned as a little kid, but "I don't understand why blah blah blah". If you like the explanation you get, the person giving it will feel good about being smarter than you, and if you don't like it, the other guy started the technical debate not you, and you never accused him/her of being "wrong". | |
| Jan 16, 2017 at 21:24 | history | tweeted | twitter.com/StackAcademia/status/821105842602778624 | ||
| Jan 16, 2017 at 20:56 | answer | added | Peteris | timeline score: 9 | |
| Jan 16, 2017 at 20:52 | answer | added | Jeff | timeline score: 28 | |
| Jan 16, 2017 at 20:34 | answer | added | Dirk | timeline score: 43 | |
| Jan 16, 2017 at 20:29 | answer | added | Jim | timeline score: 5 | |
| Jan 16, 2017 at 19:22 | answer | added | BarbalatsDilemma | timeline score: 65 | |
| Jan 16, 2017 at 18:46 | comment | added | gerrit | Also applies when replacing "statistics" to "choice of colormaps" ;-) | |
| Jan 16, 2017 at 18:45 | comment | added | Bryan Krause♦ | Perhaps consider that it will hurt a lot less coming from a colleague rather than a reviewer. Sure, there's no guarantee that a statistically-minded reviewer will be assigned the paper, but still, it's better to improve the paper before trying to publish. If the results aren't supported by statistics, they are likely to fall apart under further scrutiny, and you wouldn't want to build a research program on a false finding. | |
| Jan 16, 2017 at 18:32 | answer | added | StrongBad | timeline score: 4 | |
| Jan 16, 2017 at 18:07 | answer | added | John Feltz | timeline score: 15 | |
| Jan 16, 2017 at 17:53 | history | asked | user2390246 | CC BY-SA 3.0 |