Preparing  your  PhD  Proposal    

Nancy M. Amato

What  is  a  PhD  Proposal  really?  

•  A  brainstorming  and  planning  process  •  A  succinct  writeup  of  your  proposed  research  goals,  strategies,  jus=fica=on,  contribu=ons  

•  A  good  =me  to  get  feedback  and  direc=on  from  experts  

•  Some=mes  a  stressful  period  of  graduate  school  

Why  write  a  PhD  Proposal?  

•  An  important,  required  step  towards  gradua=on!  

•  Focus  your  research  direc=on  – Wri=ng  can  clarify  your  thoughts  

•  Establish  contact  with  commiFee  •  Obtain  feedback  from  commiFee  •  Side  effect:  Proposal  wri=ng  experience  

The  Proposal  Process  in  1  slide  •  Select  your  commiFee  with  your  advisor  

–  Get  them  on  board  –  Get  defense  date  

•  Write  proposal  itera=ng  with  advisor:    the  proposal  will  be  a  contract  with  your  commiFee  

•  Give  proposal  to  commiFee  •  Prepare  presenta=on  of  proposal  

–  Prac=ce  several  =mes  –  Get  the  =ming  down  

•  Prepare  for  hard  ques=ons  –  reread  proposal  as  commiFee  member  •  Present  and  defend  proposal  

–  Reviseproposal  based  on  feedback    &  formally  submit  

•  Celebrate  this  big  milestone!  –  Thank  commiFee,  meet  with  advisor  (and  commiFee  members)  to  discuss  ques=ons  and  

direc=ons  

•  Move  forward  on  research!!  

How  long  should  a  proposal  be?  

•  Check  your  department  &  UniversiF  rules  – Texas  A&M:  at  most  10  pages  

•  Some  do  2  versions,  one  for  commiFee  and  one  for  university  

– Columbia:    30pp  –  no  more  – Others  may  be  shorter  

What  is  the  role  of  the  CommiFee?  •  Throughout  the  process  …  

–  Guidance  and  understanding  of  what  to  expect  –  Feedback  –  Eventually,  reference  leFers  

•  At  the  proposal  – Make  sure  you  know  what  you’re  talking  about  – Make  sure  you  know  the  previous  literature  – Make  sure  you  know  the  tough  issues  and  have  some  answers  

– Make  sure  your  plan  is  reasonable  

Selec=ng  the  PhD  CommiFee  

•  Know  your  department  &  University  rules  •  Ideally  

– PhD  advisor(s)  – Faculty  in  the  topic  area  to  provide  useful  direc=on  

– Faculty  outside  topic  to  provide  high  level  feedback  

– External  member  –  someone  in  research  area  who  could  write  a  ref  leFer  

Proposal  Presenta=on  Tips  •  AFend  others’  proposal  defenses  in  your  area  •  Thank  commiFee,  introduce  yourself  and  background  •  Prac=ce  presenta=on  many  =mes  •  Be  polite  during  interrup=ons  with  ques=ons  •  Have  someone  take  notes  of  ques=ons  •  Be  open  to  sugges=ons  •  Prepare  for  ques=ons  •  Don’t  be  afraid  to  have  no  answer  for  a  ques=on.  Ask  for  

direc=on/help  on  those  ques=ons.  •  Be  confident.  Don’t  look  to  advisor  for  answers.  

Challenges,  Frustra=ons,  Misconcep=ons  

•  “The  proposal  is  just  a  hurdle.  I  can  just  propose  ideas  off  the  top  of  my  head  now  and  then  figure  out  what  I  really  want  to  do  later.”  –  the  lazy  path  

•  “How  can  I  propose  something  when  I  don’t  know  the  details  yet?”  –  the  unknown  path  

•  “I  don’t  know  how  to  organize  the  different  parts  of  the  research  on  the  page.”  –  writer’s  block  

•  “I’m  not  ready  yet.  I  might  as  well  solve  the  problems  and  then  present  them.”  –  postpone,  postpone,  postpone….  

   

What  does  a  good  Proposal  Accomplish?  

The  proposal  should  clearly  answer:  •  What  is  the  problem  you  are  studying?    •  Why  is  it  important?    •  What  results  have  you  achieved  so  far  and  why  to  

they  maFer?    •  How  is  this  substan=ally  different  from  prior  work?    •  How  will  you  systema=cally  evaluate  your  work?  •  What  do  you  need  to  do  to  complete  your  work?    •  What  is  your  =meline?  

When  is  the  best  =me  to  write/present  a  Proposal?  

•  When  you’ve  completed  the  rest  of  your  requirements  

•  When  you  have  a  clear  idea  of  the  problem  you  want  to  solve  

•  When  you  have  some  preliminary  work  done  to  demonstrate  promise  of  your  approach  

•  When  you  have  some  no=on  of  the  major  subproblems  to  be  solved  

•  When  your  advisor  recommends  

What  to  do  before  I  start  wri=ng?  

•  Think  about  what  YOU  want  to  accomplish  •  Write  a  succinct  thesis  statement/hypothesis    

•  Discuss  your  ideas  with  others  •  Present  parts  of  the  research  at  seminars,  workshops,  PhD  workshops,  conferences    

•  Think  about  3-­‐4  major  contribu=ons/papers  

•  Formulate  these  contribu=ons  in  wri=ng  

How  do  I  get  started  wri=ng?  

•  Look  at  examples  in  your  department,  with  same  advisor,  in  your  area  

•  Breadth  of  project  proposed  – Separa=on  into  subprojects  – Wri=ng  style:  problem  statement,  hypothesis  – Organiza=on  

•  Overall  outline  and  flow  • Within  each  proposed  project  sec=on  

– How  they  mo=vated  their  topic  and  foreseen  contribu=ons  

Think  about  the  Audience  •  Your  CommiFee    

–  Not  necessarily  all  in  your  general  topic  area  –  Not  familiar  with  your  specific  problem  –  Not  aware  of  your  prior  work  –  Not  aware  of  your  skills,  infrastructure  

•  Implica=ons  –  Background:  terminology,  problem,…  –  State  of  the  art  related  to  your  problem  –  Convincing  mo=va=on  for  importance  –  Demonstra=on  of  feasibility/promise  of  success  –  Is  a  star=ng  point  for  your  disserta=on/thesis  

How  to  organize  a  proposal?  •  Introduc=on    

– Problem  statement  and  importance  

•  Background  and  State  of  the  Art  •  Proposed  Research  

– Subsec=ons  on  each  research  contribu=on  •  Evalua=on  Plans  •  Research  Plan  •  Summary  of  your  Contribu=ons  and  Timeline  

The  Introduc=on    

•  General,  high  level  problem  for  people  outside  area  to  appreciate  

•  Quick  overview  of  what  state  of  the  art  does  not  address  

•  Thesis  statement  –  specific  open  problem  and  proposed  strategy  

•  Brief  overview  of  key  insights  and  why  your  approach  is  promising  

•  List  of  your  likely  contribu=ons    

Proposed  Research  

•  Overview  of  project  –  maybe  a  figure  •  Specific  project  in  steps  •  For  each  –  

– Problem  – Strategy  – Details  known  now  – Plans  for  remaining  challenges  – Evalua=on  plan  

Evalua=on  Plans  •  Experiment  Design:  

–  Ques=ons  you  will  ask  to  judge  success  of  your  approach  –  Independent  variables  –  what  is  being  varied/compared  

•  Eg,  your  technique  versus  other  techniques  –  Dependent  variables  and  measures  –  what  is  being  measured  

•  Effec=veness  –  precision  and  recall,  f  measure  •  Cost  –  efficiency  

–  Human  subjects?  – Methodology  

•  What  ac=ons  are  you  going  to  perform  to  conduct  the  experiment?  

Research  Plan  

• What  steps  do  you  plan  to  take  next?    • What  will  you  save  for  post-­‐thesis  work?    Why?  

Conclusion  

•  Summary  of  contribu=ons  to  the  state  of  the  art  –  intellectual  merit  

•  Repe==on  of  broader  impact  on  society  

Sample  Ques=ons  at  Proposal  Defence  

–  Topic  too  large?  too  small?  doable  in  the  =meframe?  Focused?  What  problems  haven’t  you  foreseen?    What  happens  if  your  planned  experiments  fail  (backup  plans)?  

–   Evalua=on  (plan,  sta=s=cs,  validity)‏  –   Related  work  missing?  –   Prac=cality/scalability  –   Vision  of  where  this  can  go…  –  Certain  people  have  go-­‐to  ques=ons  

•  Watch  your  commiFee  members  on  other  proposal  defenses  •  Especially  if  your  thesis  touches  on  work  they  have  done  or  know  a  lot  about  

•  Ask  your  advisor