Phdwiki - User contributions [en]

How to ask good questions in workshop

2009-02-17T16:29:09Z

GreenJe:

At most universities, professors from other schools will visit and present research they are currently working on. In these sessions, the audience can ask questions and make suggestions. Although not all participants may agree, the primary purpose of the workshop is to provide feedback that will help the author(s) improve the paper and potentially prepare it for publication. Keeping the purpose of the workshop in mind will help provide context and mold the content of the comments you might make. It may be helpful to consider the following questions that often come up in the referee process for journal publication:

* What is the research question?
* Is this question important in the context of the current body of research?
* Do the hypotheses effectively test the research question?
* Is the hypothesis development internally consistent?
* Can the hypotheses be rejected?
* Can the results of the hypothesis testing address the research question?
* What if the paper rejects the hypotheses, fails to reject, or goes the other way?
* Does the sample and methodology effectively test the hypotheses?

Here are a few suggestions for making workshops productive.

* Be prepared. For several reasons, in most workshops, there ''is'' a such thing as a stupid question or comment. Although you should not be excessively concerned about this, adequate preparation by reading the paper, and potentially related papers will help you distinguish between "good" and "bad" comments and make the workshop productive for you and the presenter.

* Do not ask questions or make suggestions to demonstrate your intelligence. You are there to help improve the paper not show off how smart you are.

* Seek to be constructive. Even if you find a fatal flaw in the paper, provide some suggestion for how to fix it (even if it means doing the experiment/archival analysis over again).

* Don't take your questions or the presenters response personally. It is hard to present and be politically correct at all times. Understand that the presenter is under a lot of stress and is trying to do their best to answer questions.

* Don't beat an issue to death. If a few people have made comments about a problem in the paper and the presenter has discussed it, move on to another area. Workshops get bogged down and waste the presenter's time if you stay on the same issue for the entire workshop.

* Not all research is connected to what you do. Be careful not to always tie everything everyone else does to your research (e.g., I'm interested in banking and so every paper has something to do with banking or should have something to do with banking).

* Realize that each research methodology has limitations. Be careful not to spend to much time criticizing inherent limitations to a research methodology.

* Be open minded. If everyone does the same type of research, research would get very boring very quickly. Try and see what are the benefits as well as the costs to research design choices.

* It's nice to provide written comments or marked up manuscripts to the presenter.

* Providing positive feedback can be very helpful as well. While it may not be appropriate to spend an inordinate amount of time on positive aspects of the research in workshop, a brief comment is nice to hear as a presenter. Another excellent place to provide positive comments is in written comments.

How to ask good questions in workshop

2009-02-17T16:26:52Z

GreenJe:

At most universities, professors from other schools will visit and present research they are currently working on. In these sessions, the audience can ask questions and make suggestions. Although not all participants may agree, the primary purpose of the workshop is to provide feedback that will help the author(s) improve the paper and potentially prepare it for publication. Keeping the purpose of the workshop in mind will help provide context and mold the content of the comments you might make. It may be helpful to consider the following questions that often come up in the referee process for journal publication:

* What is the research question?
* Is this question important in the context of the current body of research?
* Do the hypotheses effectively test the research question?
* Is the hypothesis development internally consistent?
* Can the hypotheses be rejected?
* Can the results of the hypothesis testing address the research question?
* What if the paper rejects the hypotheses, fails to reject, or suggests the opposite direction than what is hypothesized?
* Does the sample and methodology effectively test the hypotheses?

Here are a few suggestions for making workshops productive.

* Be prepared. For several reasons, in most workshops, there ''is'' a such thing as a stupid question or comment. Although you should not be excessively concerned about this, adequate preparation by reading the paper, and potentially related papers will help you distinguish between "good" and "bad" comments and make the workshop productive for you and the presenter.

* Do not ask questions or make suggestions to demonstrate your intelligence. You are there to help improve the paper not show off how smart you are.

* Seek to be constructive. Even if you find a fatal flaw in the paper, provide some suggestion for how to fix it (even if it means doing the experiment/archival analysis over again).

* Don't take your questions or the presenters response personally. It is hard to present and be politically correct at all times. Understand that the presenter is under a lot of stress and is trying to do their best to answer questions.

* Don't beat an issue to death. If a few people have made comments about a problem in the paper and the presenter has discussed it, move on to another area. Workshops get bogged down and waste the presenter's time if you stay on the same issue for the entire workshop.

* Not all research is connected to what you do. Be careful not to always tie everything everyone else does to your research (e.g., I'm interested in banking and so every paper has something to do with banking or should have something to do with banking).

* Realize that each research methodology has limitations. Be careful not to spend to much time criticizing inherent limitations to a research methodology.

* Be open minded. If everyone does the same type of research, research would get very boring very quickly. Try and see what are the benefits as well as the costs to research design choices.

* It's nice to provide written comments or marked up manuscripts to the presenter.

* Providing positive feedback can be very helpful as well. While it may not be appropriate to spend an inordinate amount of time on positive aspects of the research in workshop, a brief comment is nice to hear as a presenter. Another excellent place to provide positive comments is in written comments.

SAS Coding Help

2008-03-11T16:08:19Z

GreenJe: /* Replications */

== Code ==
* [http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1015021 How to Gather Data Using a Web Crawler: An Application Using SAS to Search Edgar] - interesting macro that can crawl Edgar reports for data.
* [[Basic Code]]
* [[Statistical Analysis]]
* [[Proc SQL]]

== Macros ==
* [http://www.psych.yorku.ca/lab/sas/macro.htm The basics of macros]
* [[Media:SAS_Macro_Quintiles.txt|Macro for grouping continuous variable into quintiles]]
* [[Media:WinsorizeMacro.txt|Winsorizing and trimming macro]]

==Replications==
* [[Media:BMNreplicate2.txt|Replication of Beaver, McNichols, Nelson RAS 2007]] "An Alternative Interpretation of the Discontinuity in Earnings Distributions"
* [[Media:DEAbyindustry.txt|A data envelopment analysis program]]
[[Image:SimERETOperLev2.txt]]

== Websites to Help ==
* [http://www.ats.ucla.edu/stat/stata/whatstat/default.htm What statistical test should I use]. In addition to explaining what test to use, page also provides information on how to perform test and interpret results using SAS, STATA, and SPSS software.
* [http://www.bus.umich.edu/Academics/Departments/Accounting/nhafzall/default.htm University of Michigan SAS Site]
* Another University of Michigan [http://www-personal.umich.edu/~stoffman/ SAS Site].
* Great resource for simple, straight forward explanation. [http://javeeh.net/sasintro/intro1.html Introduction to SAS: Index]
* [http://www.utexas.edu/cc/docs/stat53.html Factor analysis in SAS]
* [http://www.ats.ucla.edu/stat/examples/default.htm UCLA SAS Textbook Examples]. This page includes examples and problems worked in SAS from more than 40 statistics textbooks.

File:SimERETOperLev2.txt

2008-03-11T16:06:56Z

GreenJe: A simple simulation examining the relation between earnings and returns with a lot of assumptions, just for fun

A simple simulation examining the relation between earnings and returns with a lot of assumptions, just for fun

SAS Coding Help

2008-03-11T16:03:07Z

GreenJe: /* Replications */

== Code ==
* [http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1015021 How to Gather Data Using a Web Crawler: An Application Using SAS to Search Edgar] - interesting macro that can crawl Edgar reports for data.
* [[Basic Code]]
* [[Statistical Analysis]]
* [[Proc SQL]]

== Macros ==
* [http://www.psych.yorku.ca/lab/sas/macro.htm The basics of macros]
* [[Media:SAS_Macro_Quintiles.txt|Macro for grouping continuous variable into quintiles]]
* [[Media:WinsorizeMacro.txt|Winsorizing and trimming macro]]

==Replications==
[[Image:BMNreplicate2.txt]]
[[Image:DEAbyindustry.txt]]

== Websites to Help ==
* [http://www.ats.ucla.edu/stat/stata/whatstat/default.htm What statistical test should I use]. In addition to explaining what test to use, page also provides information on how to perform test and interpret results using SAS, STATA, and SPSS software.
* [http://www.bus.umich.edu/Academics/Departments/Accounting/nhafzall/default.htm University of Michigan SAS Site]
* Another University of Michigan [http://www-personal.umich.edu/~stoffman/ SAS Site].
* Great resource for simple, straight forward explanation. [http://javeeh.net/sasintro/intro1.html Introduction to SAS: Index]
* [http://www.utexas.edu/cc/docs/stat53.html Factor analysis in SAS]
* [http://www.ats.ucla.edu/stat/examples/default.htm UCLA SAS Textbook Examples]. This page includes examples and problems worked in SAS from more than 40 statistics textbooks.

File:DEAbyindustry.txt

2008-03-11T16:01:51Z

GreenJe: A data envelopment analysis program. Again, take it for what it is worth

A data envelopment analysis program. Again, take it for what it is worth

SAS Coding Help

2008-03-11T15:57:30Z

GreenJe: /* Replications */

== Code ==
* [http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1015021 How to Gather Data Using a Web Crawler: An Application Using SAS to Search Edgar] - interesting macro that can crawl Edgar reports for data.
* [[Basic Code]]
* [[Statistical Analysis]]
* [[Proc SQL]]

== Macros ==
* [http://www.psych.yorku.ca/lab/sas/macro.htm The basics of macros]
* [[Media:SAS_Macro_Quintiles.txt|Macro for grouping continuous variable into quintiles]]
* [[Media:WinsorizeMacro.txt|Winsorizing and trimming macro]]

==Replications==
[[Image:BMNreplicate2.txt]]

== Websites to Help ==
* [http://www.ats.ucla.edu/stat/stata/whatstat/default.htm What statistical test should I use]. In addition to explaining what test to use, page also provides information on how to perform test and interpret results using SAS, STATA, and SPSS software.
* [http://www.bus.umich.edu/Academics/Departments/Accounting/nhafzall/default.htm University of Michigan SAS Site]
* Another University of Michigan [http://www-personal.umich.edu/~stoffman/ SAS Site].
* Great resource for simple, straight forward explanation. [http://javeeh.net/sasintro/intro1.html Introduction to SAS: Index]
* [http://www.utexas.edu/cc/docs/stat53.html Factor analysis in SAS]
* [http://www.ats.ucla.edu/stat/examples/default.htm UCLA SAS Textbook Examples]. This page includes examples and problems worked in SAS from more than 40 statistics textbooks.

File:BMNreplicate2.txt

2008-03-11T15:55:07Z

GreenJe: Beaver, McNichols, Nelson RAS 2007 "An Alternative Interpretation of the Discontinuity in Earnings Distributions " Take it for what it is, I just tried to get close to the reported results for some of my own interest.

Beaver, McNichols, Nelson RAS 2007 "An Alternative Interpretation of the Discontinuity in Earnings Distributions "
Take it for what it is, I just tried to get close to the reported results for some of my own interest.