Cognitive Engineering and Decision Making Technical Group

Good morning colleagues, 

I am looking for insight from practitioners with experience doing Goal-Directed Task Analysis. Specifically: how you analyzed your interview data to generate Goal-Decision-Information Requirement hierarchies. I have reviewed some of the literature using this method and have not found details on the process used for data analysis. While I have many ideas on how it could be done, and have started - I'm looking for insight that could help validate or expedite my process. For example, did you transcribe interviews, code the content as goals, decisions, etc. and then build a tree from there? Build an initial hierarchy based on your first interview and background knowledge then iterate thereafter with data from subsequent interviews? How did you deal with bias from past domain knowledge?

This is my first post in the forum, so apologies if its not in the right place and happy to take direction on that.

Thanks in advance for your support!

Catherine 

Hi Catherine,

Glad to answer your questions on the GDTA. Usually we start with the first interview and observations/background knowledge and use it to create a draft goal structure and use them to fill out the goals, decisions, and SA requirements for the parts of the tree that we can. Then we draw on each future interview to add to/restructure the goal tree as needed and fill in more of the decisions and SA requirements. The existing draft GDTA is useful for determining where there are holes where more information is needed. Since different SMES naturally focus more on different goals and decisions, this helps you to build out the GDTA as you go along. 

I am happy to answer more questions if you have them or send papers on this process. 

best regards,

Mica

mica@satechnologies.com

Good morning colleagues, 

I am looking for insight from practitioners with experience doing Goal-Directed Task Analysis. Specifically: how you analyzed your interview data to generate Goal-Decision-Information Requirement hierarchies. I have reviewed some of the literature using this method and have not found details on the process used for data analysis. While I have many ideas on how it could be done, and have started - I'm looking for insight that could help validate or expedite my process. For example, did you transcribe interviews, code the content as goals, decisions, etc. and then build a tree from there? Build an initial hierarchy based on your first interview and background knowledge then iterate thereafter with data from subsequent interviews? How did you deal with bias from past domain knowledge?

This is my first post in the forum, so apologies if its not in the right place and happy to take direction on that.

Thanks in advance for your support!

Catherine 

Thank you so much for this Mica!

Hi Catherine,

Glad to answer your questions on the GDTA. Usually we start with the first interview and observations/background knowledge and use it to create a draft goal structure and use them to fill out the goals, decisions, and SA requirements for the parts of the tree that we can. Then we draw on each future interview to add to/restructure the goal tree as needed and fill in more of the decisions and SA requirements. The existing draft GDTA is useful for determining where there are holes where more information is needed. Since different SMES naturally focus more on different goals and decisions, this helps you to build out the GDTA as you go along. 

I am happy to answer more questions if you have them or send papers on this process. 

best regards,

Mica

mica@satechnologies.com

Good morning colleagues, 

I am looking for insight from practitioners with experience doing Goal-Directed Task Analysis. Specifically: how you analyzed your interview data to generate Goal-Decision-Information Requirement hierarchies. I have reviewed some of the literature using this method and have not found details on the process used for data analysis. While I have many ideas on how it could be done, and have started - I'm looking for insight that could help validate or expedite my process. For example, did you transcribe interviews, code the content as goals, decisions, etc. and then build a tree from there? Build an initial hierarchy based on your first interview and background knowledge then iterate thereafter with data from subsequent interviews? How did you deal with bias from past domain knowledge?

This is my first post in the forum, so apologies if its not in the right place and happy to take direction on that.

Thanks in advance for your support!

Catherine 

Thank you so much for this Mica!

Hi Catherine,

Glad to answer your questions on the GDTA. Usually we start with the first interview and observations/background knowledge and use it to create a draft goal structure and use them to fill out the goals, decisions, and SA requirements for the parts of the tree that we can. Then we draw on each future interview to add to/restructure the goal tree as needed and fill in more of the decisions and SA requirements. The existing draft GDTA is useful for determining where there are holes where more information is needed. Since different SMES naturally focus more on different goals and decisions, this helps you to build out the GDTA as you go along. 

I am happy to answer more questions if you have them or send papers on this process. 

best regards,

Mica

mica@satechnologies.com

Good morning colleagues, 

I am looking for insight from practitioners with experience doing Goal-Directed Task Analysis. Specifically: how you analyzed your interview data to generate Goal-Decision-Information Requirement hierarchies. I have reviewed some of the literature using this method and have not found details on the process used for data analysis. While I have many ideas on how it could be done, and have started - I'm looking for insight that could help validate or expedite my process. For example, did you transcribe interviews, code the content as goals, decisions, etc. and then build a tree from there? Build an initial hierarchy based on your first interview and background knowledge then iterate thereafter with data from subsequent interviews? How did you deal with bias from past domain knowledge?

This is my first post in the forum, so apologies if its not in the right place and happy to take direction on that.

Thanks in advance for your support!

Catherine