20 pager

20 pager (and submit it to the AAPI Nexus)

• Draft Deadline: July 31
• Summer/Fall 2008 AAPI Nexus
• Requirement: Manuscript length: Articles must not exceed 8000 words (including notes) or 20 pages (double spaced) inclusive of charts, graphs, notes and references.

Format: University of Chicago Manual of Style, 14th Edition.

More results and discussion

1 demographic table, 1 sub-analysis on AAPIs

(regression analysis controlling for the demographic differences, such as age and gender)

2nd wave submission for confidential data

• Errors in previous saved file -- no format saved --> use original file
  

Why is observational studies still necessary?

http://www.bmj.com/cgi/content/full/bmj%3B312/7040/1215?eaf

Randomized controlled trials

Zelen, Wennberg, Intent to treat design

Bias

Bias and causal associations in observational research

Reliability and Validity

http://webs.wofford.edu/reidak/Psych%20200/Reliability%20and%20validity.pdf

Process of writing the manuscript

• Word Template (done)
  
• Formats: Fonts, Margin, Headings
• Outlines
  
• Research methods
• Subsections:
• Read relevant examples
• Literature review
  

Research to Policy

Key words: research will inform the policy makers, transformational research

Objectives from last meeting

• Complete the application
• Human subject
• CV (short and long)
• Variable list
• Updated research description
• Analysis plan
  
• More literature /significance
• Application for the grant
• Prospectus
  

Hi Diana,

I'll re-write the application form and research proposal to incorporate
these changes and also more literature reviews. It might take a couple of
days, so I'm not going to send you the documents today.
The next review committee will be held 2 weeks from today.
I think we can take our time to catch that one.

I'll try to organize our analytic strategies. The attached is the document
from yesterday.

Don

> I think they want you to check off the needed variables in the table on
> the
> DAC website -- I think I have that... Yes. I've attached it. The yellow
> highlighting is mine, to indicated suppressed/confidential variables. I
> think you should be able to do this pretty easily from the list of
> variables
> you gave me today.
>
> Bummer about the not getting a file. Maybe we should strategize about
> which
> analyses you should do first so that (1) we're not held up and (2) when/if
> we get to give them code to run, we have a good idea of all the stuff we
> want. Can you send me the variable/analysis list you gave me today
> electronically? Or, better yet, outline the analysis steps you're going to
> take in what order -- e.g.,:
>
>         1. Descriptive statistics of all variables
>                 - (list the variables)
>
>         2. Bivariate tables of all above with the following column/outcome
> variables:
>                 - Kessler score
>                 - Self reported need for MH services
>                 - (list the rest of your desired dependent / outcome
> variables here)
>                 - (state what you will be looking for in these tables? E.g.,
> confounders, collinear variables, time dependent variables, funky sample
> sizes, etc.?)
>
>
>         3. Multivariate analyses with outcome variables as in #2
>                 - (for each outcome, list the independent / covariates you
> anticipate being in the model; this may or may not be the same for all
> outcomes)

DAC application
Description of the research.doc 

DAC datasets

• Question 1: Are they remotely available?
• Source files at UCLA Data Access Center (the “DAC”)
• Secure site at UCLA Center for Health Policy Research
• All CHIS data items except direct identifiers available, including:
• Detailed demographics
• Small geographic identifiers
• Highly sensitive items (e.g., sexual orientation, ER visits due to mental health)
• On-site or remote access available
• More than 80 researchers have used or are using the DAC
• Question 2: Are the data structure the same as the PUF or AskCHIS?

• AskCHIS uses design information in the inference while PUF is supported by the replicate weights.

• 2008 Data Users Conference Postcard - N C H S

Evidence-based Public Health

Evidence-Based Practice for Public Health
http://library.umassmed.edu/ebpph/

• Pre-Formulated and Filtered Searches of Published Studies

• Healthy People 2010 Information Access Project
• Mental Health and Mental Disorders
• Increase the proportion of adults with mental disorders who receive treatment. (Objective 18-09)
• Cochrane
  

AAPI MRP grant

• Joel as the director of CFCH has previous experience and expertise in AAPI related issues
• GSR in the DrPH program has a focus of translational research
• Previous work and extended focus, need for funding data use
• Mental health as critical issue and the policy brief would bring attention the the neglected issue