Writing Research Proposals by siy89361

VIEWS: 5 PAGES: 27

									  Writing Research Proposals

Margarita Alegria, Professor Harvard
          Medical School,
    Cambridge Health Alliance
          Academy Health
           June 4, 2007
I. Overviews of research branches and support
   mechanism

II. Formulations of research problem
     A. Good idea of research priorities of funding
   institute
         1. Concepts paper, pre-proposal
     B. Get acquainted with literature review
     C. Contact with researchers in the area

III. Thinking about theory/concept: development of a
   theoretical/conceptual framework and rationale for
   research proposal

IV. Selecting critical issues
    A. Narrowing the topic
    B. Selecting the target population
    C. Aims and preliminary studies
Databases with NIH Information
   CRISP (Computer retrieval of information on
    Scientific Projects)

       Contains information on research projects and
        programs funded by DHHS including NIH, CDC,
        FDC, HRSA and AHRQ
        Useful for identifying the appropriate institute
        to fund a grant, the best study section to
        review it, and the most appropriate funding
        mechanism.

   NIH Guide for Grants and Contracts

       contains weekly announcements detailing
        extramural funding programs and policies of
        the NIH and other public service entities.
       Do Some Homework
   According to studies, as many as 80% of all applications
    are misdirected or inappropriate.

   Become familiar with a source before applying. If a phone
    number or email address is provided, call and discuss
    whether your work matches their interests.

   The Foundation Center is a major publisher of non-federal
    funding opportunities and has current and historical
    information on every foundation in the U.S.
I.        Review Criteria
     A.    Scientific significance of idea
     B.    Originality/Innovation
     C.    Rigor of design and
           methodology
     D.    Data analysis strategy
     E.    Human subjects
     F.    Research environment
     G.    Investigative team
     H.    Budget
Review Criteria and rating of unsolicited
research grant and other publications

   Significance
      Does this study address an important problem?
      How will scientific knowledge be advanced and what effect
       will the study have on the concepts or methods that drive
       this field? More and more seen in applied terms

   Approach:
      Are the conceptual framework, design, methods, and
       analyses adequately developed, well-integrated, and
       appropriate to project aims?
      Does the applicant acknowledge potential problem areas
       and consider alternative tactics?

   Innovation:
      Is the project original, employing novel concepts,
       approaches or methods?
      Does the project challenge existing paradigms or develop
       new methodologies or technologies?
     Developing the Research Question
   Statement of the problem
       What do you want to do?
       What questions are you asking in the research?
       What are your goals?
   State your objectives clearly. Be specific and
    concrete (use measurable concepts).
   State your objectives in declarative form.
   List your objectives in order of priority or as
    building blocks and follow this order
    throughout proposal (i.e., use parallel
    structure).
     Example of Specific Aims of R01
      The central aims of NLAAS are:
1. To estimate the lifetime and 12-month
   prevalence of psychiatric disorders and the
   rates of mental health services use for
   Latino and Asian American populations,
   adjusting for age and gender effects using
   nationwide representative samples of
   Latinos and Asian Americans.
2. To estimate the relation of social position,
   environmental context, and psychosocial
   factors with the prevalence of psychiatric
   disorders and utilization rates of mental
   health services in nationwide
   representative samples of Latinos and
   Asian Americans.
Example of the Research Problem


 Project Objectives: We will compare the cost,
 sustainability and effectiveness of 3 innovative
 strategies to improve the treatment of depression
 in primary care. The three strategies are: 1)
 financial incentives for providers; 2) a program
 designed to help patients communicate more
 effectively with their physicians regarding their
 depression; 3) the use of volunteer aides who act
 as peer support counselors and advocates with
 the care team .
    Significance of Research Problem
   Why are your questions and answers
    important?
   Theoretical significance: contribution to
    basic knowledge
   Methodological significance: new
    methods will be developed or old methods
    adapted
   Applied significance: solution of a specific
    real-world problem
   Social significance: benefits to society at
    large
Example of Significance of Research Problem


Considerable evidence exists in the clinical research
literature supporting the view that Attention-Deficit
/Hyperactivity Disorder (ADHD) is a disabling disorder that
affects individuals across the lifespan (Barkley, 2003).
However, it is not clear whether the correlates,
comorbidity, and impairment of the disorder in probability
treatment referral samples are similar to those identified
in probability community samples. These variables have
not been systematically compared in the two populations
using the same methods. This comparison is important
since findings from convenience or probability treatment
samples can be consonant with clinician’s experience but
biased and unrepresentative of the general population
with the disorder. This discrepancy was coined by Cohen
and Cohen (1984) as “the clinician’s illusion” i.e., the
attribution of the characteristics of patients with a disorder
identified through clinical experience (one could also add,
through non probabilistic treatment samples) to the entire
population with the disorder.
      Review of Relevant Literature
   How have these questions been
    answered before? [Have the questions
    been asked? Have related questions
    been asked (e.g., with different
    populations)? What answers have been
    obtained?
   What are the limitations of past
    research?
   How will you build on past strengths
    while overcoming limitations? Why is
    your research worthwhile and necessary?
   Your review of previous research
    should be critical and evaluative
    synthesizing rather than
    comprehensive. Show that you know
    the past research so well that you can
    summarize its major themes, findings,
    strengths, and weaknesses.

   Maintain parallel structure in this
    section. Discuss the literature relevant
    to your first aim/question (and
    hypothesis) first, then discuss the
    literature for your second
    aim/question, and so on.
     Hypothesis
   What answers are you proposing for the
    research question?
   A hypothesis describes a relationship
    between two or more variables.
   In your hypotheses, the variables must
    be identifiable conceptually and capable
    of empirical observation.
   Similarly, the relationship among the
    variables (and changes in those
    relationships) must be identifiable
    conceptually and operationalizable.
   Your hypotheses must be testable.
   Order your hypotheses to correspond to
    your already-stated research questions.
   Develop alternative hypotheses whenever
    possible. In predicting a particular
    relationship between variables (or
    explanation for a particular variable),
    consider alternative explanations and test
    for them in your research. Anticipate
    alternative explanations for your expected
    results and devise additional tests to falsify
    those competing hypotheses.
                                         Conceptual Framework



Background factors           T1 Stress                 T2 Mediators         T2 Outcomes


                                                       T2 Mediator          T2 Mediator        T3 Outcomes



       Age           Individual                    Social support          Drug Use          Problem behaviors
                       Psychological distress       Family support          Use/No use
       SES             Physical illness             Friend support          Severity      Criminal behavior
                       Attitudes towards                                    Frequency      Property crimes
   Marital Status    deviance                      Institutional Support                   Drug dealing
                       History of conduct            Human service                         Prostitution
    Employment       problems                      programs
                                                     Drug treatment                       High Risk Sexual Activity
     Education       Social                                                                Condom Use
                      Family/Friends attitudes                                             Number of partners
                                                       Incarceration
                     and modeling of drugs
                      Role/Life stressors                                                 Marital Violence
                      Perceived educational
                     and job opportunities                                                Health Outcomes
                      Childhood victimization                                              HIV/STDs
                                                                                           Functional Impairment
                     Cultural
                       Neighborhood norms and
                     trends
                       Family affiliative values
Hypotheses of Remission From Drug Dependence Symptoms and
Drug Use Cessation


   Overall, low socioeconomic               Relationships with drug
    status, comorbid                          using partners are
    psychiatric conditions, and               strongly associated with
    lack of family and social                 relapse of drug use (Tuten &
    supports are consistent                   Jones 2003) and post-treatment drug
                                              use continuation (Amaro & Hardy-
    markers of non-                           Fanta 1995; Grella, Scott, Foss, Joshi,
    compliance with                           & Hser, 2003).
    treatment, and relapse
    following treatment
    completion (Daley & Marlatt
    1992).                                   Moreover, childhood
   For women in particular,                  experience of sexual
    symptoms of depression                    abuse is a significant
    are believed to play an                   predictor of drug use
    important role in                         disorder, as well as a risk
    persistence of substance                  factor for unsuccessful
    abuse (Galaif, Nyamathi, & Stein,         treatment and eventual
    1999; Wang, Collins, DiClemente,          relapse (Snow & Anderson 2000).
    Wingood, & Kohler, 1997; Zuckerman,
    Amaro, Bauchner, & Cabral, 1989).
Study Design
  1.   Are the design and measurement strategies
       the best ones for testing the hypotheses in
       terms of sensitivity to processes under study,
       experimental control, and statistical power?
       Are independent and depend variables and
       covariates clearly specified?
  2.   Has any preliminary (pilot) work been carried
       out to help document feasibility of recruitment
       and methods, and the validity of the
       measures?
  3.   Is sample adequately described? Is sample
       representative of population or subpopulations
       of interest? Has subject attrition been
       considered? Are criteria for inclusion/exclusion
       clear and complete?
4.    Are the manipulations of independent
      variables potent enough to result in
      changes in outcome (dependent) variables?
5.   If some of measures are based on self-
      report, are there other measures that
      would help to establish their external
      validity?
6.   Have all possible confounds been
      considered, such that the results will test
      the effects of the manipulation or process
      of interest?
7.   Is the study replicable in terms of sampling
      strategies and measurement methods?
8.   Are results likely to be generalizable, and
      what conclusions will be drawn?
9. Have the investigators considered
   limitations and potential problems, and do
   they indicate how they will deal with these
   problems? Have sources of bias been
   considered (testers blind to group
   placement, recall response bias)? If testing
   is repeated, are measures susceptible to
   practice effects or reactivity?
10. Does the proposal reflect a process of
   “thinking through” the issues under
   investigation from the available research
   literature to the testable hypotheses,
   operational measurement, expected
   results, and interpretation of alternative
   findings?
      Study Design of Inner City Latina Drug
      Use Study (ICLDUS)
Three-wave longitudinal research project to study drug use
  patterns and correlates among low-income inner city
  Latina women, ages 18 to 35. Women were interviewed at
  baseline (Wave I) and at two follow-ups (Wave II and III)
  with an average of fourteen months between data
  collection periods. Two groups of women were studied: a
  sample of 275 crack/cocaine or heroin drug-using women
  and a sample of 443 community women living in the same
  neighborhoods, extended metro San Juan area.

Two groups of women were recruited for the study:
  cocaine/crack and injecting drug users from drug copping
  areas (e.g., crack houses or shooting galleries where
  drugs are sold), and non-drug users who lived within a
  one-mile radius of the copping areas.
       Study Design-Continuation
   ICLDUS survey instrument was administered in 1996-1997
    to a sample of 718 women, for a response rate of 81.4%.
    The first follow up was conducted with a total of 696
    women between July 1998 and June of 1999. A total of
    681 women were re-interviewed for the second follow up
    during July of 1999 through November of 2000.

   Respondents who remained in the study for the three
    waves as compared to those who dropped out from Wave
    I to Wave II (n=54) or from Wave II to Wave III (n=29)
    were more likely to be from the copping sample, older
    (30-38 years of age), returned migrants to the island,
    self-reported drug users, HIV positive, and women
    practicing sex work.
Outcome Measures

Last year arrest -established by asking respondent if they
  had been arrested (yes=1) within last 12 months.
Practicing sex work by query asking if respondent had
  engage in any sexual practices in last 12 months in
  exchange for money, drugs or other personal favors; (1)
  if yes, 0 if otherwise.
An item asked if they had committed any suicide attempt
  in the last year, which was coded 1 for those who said
  yes, 0 if otherwise.
Intimate partner violence- recount of times, during the
  last 12 months, their partner/spouse responded to
  domestic conflict through inflicting physical of
  psychological harm using the Conflict Tactic Scale (Straus
  1990a; Straus 1990b).

Death was assessed by reports of death of the respondent
  and later confirmed with death certificate or by
  testimony of family and friends.
      Study Design of System Cost
      Effectiveness and Managed Care


The objective of study is to evaluate the cost
effectiveness of managed care as compared to
public health system care on the treatment of
depression. Tracking depression Tx allows us
to assess system effects in the distribution of
resources to untreated, ineffectively treated
and effectively treated cases proportional to
costs. Is managed care able to provide more
effective depression Tx for the same costs as
the public health system?
  Definition of Appropriate Treatment
• No Treatment:
      Persons who did not receive any formal mental health
      treatment (no visits w/in a 12 month period to mental
      or general health provider to discuss mh problem).
• Not Effective Treatment:
      Persons who received minimal counseling (less than 4
      visits) or those without antidepressant medication, or
      those that report no visits but do report receiving
      antidepressants under no medical supervision.
  • Effective Treatment:
      Persons who received effective counseling (4+) with
      no effective medication or received minimal or
      effective counseling (1-3 visits) and antidepressants.
V. Ten (10) hints
    A. Read instructions
    B. Follow page limitations – make use of appendices
    C. Use as an example another successful proposal in
   the area of interest
    D. Develop a team – if you do not have experts, get
   consultants
    E. Don’t assume reviewers will know – be very
   specific (if no pilot data, you are in trouble)
    F. Clear concise writing style
       a. Keep together related ideas
       b. Shorten long sentences
       c. Eliminate redundant ideas
    G. Allow time for thorough editing – send your
   proposal to experts (try to get critical input upfront rather
   than back end)-Do not submit a proposal for consultation
  H. Include potential pitfalls and limitations
  I. Contact member of program review for advice
  J. The situation for new applicants is difficult, but not
  impossible

VI. Fatal mistakes in grant writing
     A. Let’s get a grant to pay for treatment or services
     B. Signing up for the wrong race (do what you do
  best)
     C. Trust me - I’m an expert (skipping the details or
  concrete information)
     D. Ignoring the pink sheet
     E. I’m going to develop ….
     F. Excuses – explaining why the study can’t be
  done the right way
    G. Not re-submitting
    H. Being out of sync w the field-only found through
  networking (eg. one more study in depression in
  primary care)

								
To top