Moving from Excellent (2.0) to Outstanding (1.0) by tum19250

VIEWS: 9 PAGES: 20

									Moving from Excellent (2.0) to
     Outstanding (1.0)
    Confessions and Obsessions of a
      Member of a Study Section



               Judith M. Ford
            Psychiatry Department
      Yale University School of Medicine
My own perspective as a reviewer
 • Reviewing grant proposals is difficult and
   extremely time-consuming.
 • Your proposal will be read by 3 people, some
   of whom may be only relatively familiar with
   your topic.
 • They are your advocates before the committee.
    – Make them like you.
    – Help them understand what you propose.
    – Don’t annoy them with inconsistencies, typos,
      errors, arrogance, narrow margins and a small font.
                  Before you start to write
        • Pick a topic that you can “pitch” in 30
          seconds or less.
        • The primary reviewer of your proposal
          has only a few minutes to make that same
          pitch to the committee.
        • The primary reviewer of your proposal
          might say something like this:
“This is a proposal by an established team to test the hypothesis that a dysfunction in the corollary
discharge mechanism underlies auditory hallucinations. The corollary discharge is a mechanism
by which sensory perception is reduced when sensation results from self-generated actions. The PI
proposes to collect ERP and fMRI data while subjects hear speech that they produce as they speak,
that they deliver themselves by pressing a button, and that follows a visual cue. The PI has
published data showing that sensation is reduced to self-produced speech in normal controls, but
not in patients with schizophrenia. Thy propose to also collect DTI data to allow assessment of
fiber tract integrity between speech production and perception areas of the brain. As such, this is
an important extension of the previous work………”
            Title and Abstract
• It needs to communicate clearly what you propose
  to do.
• It may determine which study section it will go to.
• It may determine who is asked to review your
  proposal.
• Reviewers will rely heavily on the abstract when
  providing a summary in front of the committee.
           Specific Aims
• This is the first section you should write.

• Include a brief introductory paragraph to lay
out the rationale.

• State your hypotheses or questions clearly.

• Be sure they are testable and that you can test
them.

• Keep it short--about one page in length.
                                      Specific Aims
Specific Aim 1. Basic Explorations of Corollary Discharge (CD) Dysfunction
   a. Congruence Between Speech Acts and Sensory Consequences
              To refine our neurobiologic observations of CD deficits in
              schizophrenia we ask: Does suppressive action of the CD during
              self-produced speech (talking ) depend on a tigh t congruence
              between properties of the motor act and its sensory consequ ences?
              Is CD suppression dependent on sensory expec tancy rather than
              motor-sensory congruen ce per se? Does the CD adap t with
              repetition of distortion and delay?
   b. Speech Acts vs. Simple Motor Acts
              The link between a motor speech ac t and the sound produced is                         Note wide margins,
              over-learned. But perhaps motor acts in general temporarily
              suppress resulting sensory inputs. Does pressing a button to
              deliver a tone send an "efference copy" of the expected tone to the
                                                                                                     and bolded Aims.
              auditory cortex reducing its responsivity to it? Is the CD deficit
              seen in schizophrenia specific to over-learned speech ac ts and the
              resulting phon eme or to a general deficit in expectancy?
                                                                                                     This makes it easy to
   c. Meaningful Words vs. Non-meaningful Speech Sounds
              Are CD effects greater to meaningful speech than to non-
              meaningful utterances, and is CD dysfunction to meaningfu l
                                                                                                     read, and leaves room
              speech even greater in patients? Is frontal-temporal connec tivity
              during word generation and non-meaningful utterances simil ar?
   d. Temporal Course of CD Suppression
                                                                                                     for positive comments
              Are CD deficits in schizophrenia related to deficits in the timing of
              communication between frontal and temporal areas during talking?                       in the margins.
              What is the CD mechanism’s time course with respect to the
              initiation and execu tion of speech?

Specific Aim 2. Relating Integrity of White Matter Tracts Connecting Fro ntal and
Temporal Lobes to ERP and fMRI Measures of CD Dysfunction
              Is cortical suppression during talking directly linked to integrity of
              white matter tracts linking speech production and speech re ception
              areas, and not to the integrity of other fiber tracts?

Specific Aim 3. Relating Corollary Discharge Dysfunction to Hallucinatory State and Trait
              Do ERP and EEG coherence indices of CD failure remit following
              rTMS treatment for auditory hallucinations?
  CD = corollary dis charge                    DLPFC = dorso-lateral prefrontal cortex
  ERP = event related potential                ACC = anterior cingulate cortex
  fMRI = functional magnetic resonance image   rTMS = repetitive transcranial magnetic stimulation
  DTI = diffusion tensor image                 DA = dopamine
  MEG = magnetoencephalography                 BA = Brodmann’s Area
  Glossary--if you use abbreviations,
please, please, please include a glossary
       Specific Aim 1. Basic Explorations of Corollary Discharge (CD) Dysfunction
          a. Congruence Between Speech Acts and Sensory Consequences
                     To refine our neurobiologic observations of CD deficits in
                     schizophrenia we ask: Does suppressive action of the CD during
                     self-produced speech (talking ) depend on a tigh t congruence
                     between properties of the motor act and its sensory consequ ences?
                     Is CD suppression dependent on sensory expec tancy rather than
                     motor-sensory congruen ce per se? Does the CD adap t with
                     repetition of distortion and delay?
          b. Speech Acts vs. Simple Motor Acts
                     The link between a motor speech ac t and the sound produced is
                     over-learned. But perhaps motor acts in general temporarily
                     suppress resulting sensory inputs. Does pressing a button to
                     deliver a tone send an "efference copy" of the expected tone to the
                     auditory cortex reducing its responsivity to it? Is the CD deficit
                     seen in schizophrenia specific to over-learned speech ac ts and the
                     resulting phon eme or to a general deficit in expectancy?
          c. Meaningful Words vs. Non-meaningful Speech Sounds
                     Are CD effects greater to meaningful speech than to non-
                     meaningful utterances, and is CD dysfunction to meaningfu l
                     speech even greater in patients? Is frontal-temporal connec tivity
                     during word generation and non-meaningful utterances simil ar?
          d. Temporal Course of CD Suppression
                     Are CD deficits in schizophrenia related to deficits in the timing of
                     communication between frontal and temporal areas during talking?
                     What is the CD mechanism’s time course with respect to the
                     initiation and execu tion of speech?

       Specific Aim 2. Relating Integrity of White Matter Tracts Connecting Fro ntal and
       Temporal Lobes to ERP and fMRI Measures of CD Dysfunction
                     Is cortical suppression during talking directly linked to integrity of
                     white matter tracts linking speech production and speech re ception
                     areas, and not to the integrity of other fiber tracts?

       Specific Aim 3. Relating Corollary Discharge Dysfunction to Hallucinatory State and Trait
                     Do ERP and EEG coherence indices of CD failure remit following
                     rTMS treatment for auditory hallucinations?
         CD = corollary dis charge                    DLPFC = dorso-lateral prefrontal cortex
         ERP = event related potential                ACC = anterior cingulate cortex
         fMRI = functional magnetic resonance image   rTMS = repetitive transcranial magnetic stimulation
         DTI = diffusion tensor image                 DA = dopamine
         MEG = magnetoencephalography                 BA = Brodmann’s Area
                       Background
• Be clear and scholarly.
• Give road signs to help the reviewer follow your argument.
        If you can, follow organization of Specific Aims.
• Review relevant literature.
        Don’t paste in your dissertation.
• It is helpful for citations in text to give author’s name (not a
   reference number).
• In Reference Section, include all authors of the paper.
        It helps to know the lab where the work was done.
• Make sure you are citing the literature accurately.
        You lose points if you don’t.
• Make sure you include recent studies in review.
• At the end, summarize what is known and where the gaps in our
   knowledge are.
               Preliminary Studies
• Pilot data are essential for an R01(but not for R21 or R03).
• Pilot data demonstrate feasibility of your method and show your
  hypotheses are promising.
• Show figures of data you have collected.
  -be sure they can be read by the naked eye.
  -be sure they are correctly labeled and annotated.
  -be sure they are correctly called out in the text.
  -be sure your written description of the data matches the visual
  impression of the data.
  -be sure figures can be seen after your proposal has been
  scanned and made into a PDF (a degrading process).
• At the end, summarize what you learned from your pilot studies
  and what you want to do next.
                          Methods
• Give overview (again).
        Reviewers may not read the whole proposal in one sitting.
• Put in a time line.
         Make a chart to justify a 5-year study.
• Give the numbers of subjects to be studied.
• Give inclusion and exclusion criteria.
        Avoid inconsistencies in age, diagnosis, gender, etc.
• Describe methods carefully.
        The devil is in the details.
• It helps to show predictions with graphs.
        If you can’t do this, the design is too complicated.
• Provide a conceptual/theoretical rationale for your predictions.
        If you can’t do this, start over.
• At the end, relate your predictions to your specific aims and
  hypotheses.
                                Time Line
 D. RESEARCH DESIGN AND METHODS
 1. Timetable and Overall Plan
                Specific Aims 1 and 2 will be implemented at Stanford University, and Specific
                Aim 3 will be implemented at Yale University. For Specific Aim 1, we propose a
                series of ERP and fMRI studies of patients with schizophrenia and con trols to be
                conduc ted over five years. For Specific Aim 2, we will acquire DTI after fMRI
                session s to determine integrity of fronto-temporal white matter tracts. For
                Specific Aim 3, we will conduc t studies designed to assess whether corollary
                discharge dysfunction, as assessed by ERP and EEG coherence measures, is a trait
                feature of patients who have a propen sity to hallucinate, or a state feature which
                normalizes with successful rTMS treatment for hallucinations. Diagno sis, clinical
                assessment, and ERP recording will be the same at both sites.
5-YEAR TIME LINE                             Year 1 Year 2 Year 3 Year 4 Year 5
Basic Corollary Discharge Studies
Exper iment 1 (ERP, fMRI, DTI)                -------------------- --->
Exper iment 2 (ERP)                           -------------------- ---->
Exper iment 3 (ERP, fMRI, DTI)                                     ------------ ---------->
rTMS studies at Yale                                                            --------------------->
                                                ---------------------------------           -
Data analysis and interpretation                                                --------------------->
                                                       --------------------------           -



With this in hand, the reviewer can explain to the committee why you need
5 years of funding.
          Experimental Parameters
                   is individually molded using dental impression material to allow flexibility
                   in positioning the subject's head, is comfortable to use, and is effective in
                   reducing maximum head movement to less than 0.5 mm (translation) and
                   0.5o rotation. It is especially useful in minimizing head movements during
                   talking. With only the upper incisors resting in the bite bar, subjects have
                   no difficulty producing speech sound s, which are recorded using a non-
                   ferrous microphone attached to the bite bar with surgical tape. Each
                   protocol is summarized in the following table and described in more detail
                   below. Total time for each scan session, allowing for instruction,
                   changing parameters between sequences, adjusting shims, and prescribing
                   slice selection, is ~ 2 hours.
         Sequence                                     Param eters
                                                                                           o
                           Coronal localizer          TR=18ms; TE=6 ms; flip angle=45 ; FOV=24cm; 5 slices,
                                                      5mm thick
                                                                                          o
         SPGR sagittal anatomic scan                  TR=22ms; TE=3ms; flip angle=35 ; FOV=24cm; 124 slices,
                                                      1.5mm thick; reconstructed SPGR image is a 124 x 256 x
                                                      256 matrix (1.5 x 0.9 x 0.9 mm)
                                                                                             o
         Shim correction (for fMRI to correct         TR=1000 ms; TE=7ms; flip angle=60 ; FOV=32cm; 32
         for magne tic inho mogeneities)              slices, 7.5 mm thick; 0 gaps
         Functional scans (see below for              TR=6000 ms; TE=30ms; FOV=24c m; 64x64 spiral gradient-
         details )                                    echo matrix; 24 slices, 6mm thick
         Shim correction (for DTI with fewer TR=1000 ms; TE=7ms; flip angle=60 o; FOV=32cm; 12
         slices than fMRI)                            slices, 7.5 mm thick, 0 gaps
         DTI (see below for details)                  TR=8000 ms; TE=80ms; single-shot spin echo EPI technique;
                                                      FOV=26c m; 128x128 matrix; zero fill ed to 256x256 ; 12
                                                      slices, 6mm thick, 0 gap



It is easy for the reviewer to find important details if they are in a
table, and to skip them, if they are not important to the reviewer.
                        Power Calculations
              iv. Power Calculations
          We have used ex isting ERP data to derive estimates of several of the important N1 effects.
     We have used E RP data from Cz in our paper comparing the effects of Talking and Playback in
     patients and controls (45) to derive estimates of power for the effects of Group x Active/Passive
     (i.e., Talking vs. Playback ). We used data from Fz and Cz in our pilot study of pure and
     distorted speech to derive estimates of power for the effects of Talking vs. Playback x Normal
     vs. Distorted. Although we have tested only one patient in that study, we are using his mean N1
     amplit ude and the standard deviation from the controls to estimate power for the Group x
     Talking x Distortion effect. This may have given us an unrealistic estimate of our power. We
     have only tested a single control in our Tone expe riments and cannot estimate power.
     Assuming a two-group (controls vs. patients) or a three-group (controls vs. hallucinators vs.
     non-hallucinators) ANOVA model with equal variances and two-tailed significance tests with
     alpha = .05, sample sizes of 15 hallucinators and 15 non-hallucinators and 15 controls will
     enable detection of the mean differences indicated in the following table with moderate to high
     degrees of power. Estimates below are based on 15 patients vs 15 controls. We propose to use
     these power estimates to determine the number of subjects for the combined ERP/fMRI studies.
                           ANOV A Effect N1                 Mean #1       Mean #2       Pooled SD POWER
     Talking (T) vs. Playback (P) for Veridical Speech in T = -0.96       P = -3.08        2.76        84%
     NCS
     Talking vs. Playback effect for Veridical (V) vs.      V = 3.24      D = 2.63          .93        72%
     Distortion (D)
     Talking vs. Playback effect for NCS vs. SCZ            NCS = 2.13 SCZ = -0.44         2.42        98%
     Group x Talking x Distortion                           NCS = .61 SCZ = -3.3            .93       100%
     [Veridical (P-T) - Distorted (P-T)]
     5. Specific Aim 2: Relating Integrity of White Matter Tracts Connecting Frontal and
     Temporal Lobes to ERP and fMRI Measures of CD Dys function


It is difficult to get funded without power calculations that justify the sample size.
          Weeks before it is due
• Give it to someone to read.
  …Your mentor, a colleague, a friend in the English
  department, your mother. Ask for brutal criticism.
• It’s better to get negative feedback from friends than
  from the committee.
• If it is not “good to go”, wait to submit until next
  round.
• Proof-read it. Typos make the reviewer’s job harder.
                     Appendix
• The appendix is NOT part of the electronic materials that are
  sent to the reviewers.
• All essential material should be in the body of the proposal,
  within the 25 page limit.
• Do not put a glossary of terms or abbreviations in the
  appendix.
• Put in articles that will help the reviewer understand the
  detail of the project.
• You cannot expect every reviewer to read all of what you put
  in the appendix.
• Reviewers may not be required to read the appendix.
           Remember
•   Font
•   Line spacing
•   Wide Margins
•   Figures
•   Charts
•   Tables
•   Avoid inconsistencies
•   Make it easy to read.
•   Make it enjoyable to read.
•   Be sure it makes sense, all the way through.
What happens if you don’t get
         funded?
        Responding to the Pink Sheets
•     You have 3 pages to reply to the critique.
             Use every bit of it.
•     Do not continue your response in an appendix.
             That is very annoying, and it may not get read.
•     Do not simply refer the reviewer to a page in the text to find the changes you
      have made.
             Explain the gist of how you responded.
•     Respond to EVERY item in the review.
•     If you disagree, explain why.
             “We disagree” doesn’t work.
•   If the reviewer asks for more detail, do not just refer to a paper, especially if it is
      not appended.
•   Revise the body of the proposal to be consistent with the changes you said you
      would make in the reply to the review.
•   Gray-line important changes in the text.
             Do not say the changes are too extensive to denote.
•   If you made mistakes in the first submission, apologize.
     Get feedback from your project officer who heard the
     committee’s discussion of your proposal.
            Good Luck and
           Enjoy the Process
• Writing grant proposals is an important part
  of the research process.
• It focuses your thinking.
• It makes you read the literature.
• It forces you to put your ideas into action.
            I thank….
 All of you who take my advice to heart
         and make my job easier.
    Friends on my Study Section who
responded happily to my request for ideas
              for this talk.

								
To top