View previous topic :: View next topic |
Author |
Message |
Vaqueiro
Joined: 18 Jun 2008 Posts: 33 Location: Japan
|
Posted: Tue Nov 20, 2012 11:42 am Post subject: |
|
|
Glenski wrote: |
May I suggest spending some time creating something that will help you gather information anonymously and more efficiently: SurveyMonkey? |
Agreed. Survey Monkey is an excellent service. |
|
Back to top |
|
|
spiral78
Joined: 05 Apr 2004 Posts: 11534 Location: On a Short Leash
|
Posted: Tue Nov 20, 2012 3:19 pm Post subject: |
|
|
You'll have to somehow get quite a few extremely generous souls to fill out your comprehensive survey in order to ultimately have anything of value, and even then things will be in continuous flux.
Do let us know how it goes. I'll be curious when/if you get actual data.
for what it's worth, I'll suggest that you further limit the survey. It would likely be more useful to focus on
1. Jobs that don't require specialist qualifications: probably teaching general English or general EAP, at intermediate or lower levels (in my experience higher level speakers usually want in-field courses, which by definition should require at least some relevant quals/experience for the instructor).
2. Jobs in regions where openings are plentiful enough so that local connections and reputatio are not necessary. |
|
Back to top |
|
|
Sashadroogie
Joined: 17 Apr 2007 Posts: 11061 Location: Moskva, The Workers' Paradise
|
Posted: Wed Nov 21, 2012 5:01 am Post subject: |
|
|
Of course, nobody could do my job without a strong liver, hic! |
|
Back to top |
|
|
HLJHLJ
Joined: 06 Oct 2009 Posts: 1218 Location: Ecuador
|
Posted: Wed Nov 21, 2012 10:32 am Post subject: |
|
|
Vaqueiro wrote: |
I have years of experience with research methodology and statistics, and I'm considering using it to make information more accessible in our field. |
Then you should be able to come up with something significantly better than this. The data you are requesting is meaningless and close to unanalyzable. There are too many variables, most of which you aren't even recording, and none of which are constant. Sure you could collect some figures that you could cram into a stats program but the resulting answers would be as nonsensical as the data that went in. |
|
Back to top |
|
|
Vaqueiro
Joined: 18 Jun 2008 Posts: 33 Location: Japan
|
Posted: Wed Nov 21, 2012 2:52 pm Post subject: |
|
|
spiral78 wrote: |
You'll have to somehow get quite a few extremely generous souls to fill out your comprehensive survey in order to ultimately have anything of value, and even then things will be in continuous flux.
Do let us know how it goes. I'll be curious when/if you get actual data.
for what it's worth, I'll suggest that you further limit the survey. It would likely be more useful to focus on
1. Jobs that don't require specialist qualifications: probably teaching general English or general EAP, at intermediate or lower levels (in my experience higher level speakers usually want in-field courses, which by definition should require at least some relevant quals/experience for the instructor).
2. Jobs in regions where openings are plentiful enough so that local connections and reputatio are not necessary. |
Thanks for the recommendations. Of course, you were correct that this thread did not receive much attention, and I believe that I need both to set up a proper survey and to advertise it on the proper channels. Back to the drawing board... |
|
Back to top |
|
|
Vaqueiro
Joined: 18 Jun 2008 Posts: 33 Location: Japan
|
Posted: Wed Nov 21, 2012 3:13 pm Post subject: |
|
|
HLJHLJ wrote: |
Vaqueiro wrote: |
I have years of experience with research methodology and statistics, and I'm considering using it to make information more accessible in our field. |
Then you should be able to come up with something significantly better than this. The data you are requesting is meaningless and close to unanalyzable. There are too many variables, most of which you aren't even recording, and none of which are constant. Sure you could collect some figures that you could cram into a stats program but the resulting answers would be as nonsensical as the data that went in. |
I took freshman stats as well, and I have leveled those same pat criticisms at people. It comes off as intelligent as long as you aren't forced to provide solutions or concrete information.
From your comment, you seem to think I'm going to plug all of this data into a single model, which is far from the case. Some items could be compared and some would have to be viewed independently. You did not take the time provide examples of which items were "nonsensical" and which were not taken into account; this would be helpful. I'm not trying to say that my ideas are perfect. Survey generation is trial-and-error. Ideas? |
|
Back to top |
|
|
HLJHLJ
Joined: 06 Oct 2009 Posts: 1218 Location: Ecuador
|
Posted: Wed Nov 21, 2012 3:43 pm Post subject: |
|
|
Good survey design should not be trial and error and the issue isn't with your items it's that your basic premise is flawed. You have no constants and nothing that is logically or demonstrably comparable, therefore nothing meaningful to analyse.
If you are serious about doing this I will try to help, but there is nothing I can work with here. In the meantime I'll give you the same advice I'd give my students. This won't work and it's not salvageable, go back to basics and try again. |
|
Back to top |
|
|
|