Choose Language Hide Translation Bar
trish_roth
Level III

Using JMP To Manage a Learning Program To Develop the Next Generation of JMP Data Ninjas (2023-EU-30MP-1278)

Trish Roth, Principal Scientist, Abbott Laboratories
Jeff Pennoyer, Manager Environmental Monitoring and Microbiology, Abbott Laboratories

 

Many know JMP as a powerful tool for analytics and modeling and aspire to leverage JMP’s advanced capabilities to champion improvements and business understanding. It can take time and domain experience to achieve a high level of proficiency. Don’t dismay; we all start somewhere! Even at modest experience levels, value can rapidly be achieved using JMP fundamentals. Fundamentals can be quickly propagated across an organization to seed and inspire a culture of analytics. Hear how our team has integrated offerings from JMP education in a JMP “boot camp” format. The faster an organization can establish basic proficiency in JMP, the sooner it can benefit from that investment. Additionally, having a shared platform for both basic and advanced analytics creates a collaborative community, increases self-sufficiency, and provides a learning path to foster employee development. While sharing our training approach, we will demonstrate foundational JMP features, including data filters, tabulate, summary, recode, column formula, and column properties functions to track student progress. See JMP in action as we highlight methods to construct, customize and journal graph builder visuals in ways that entice spreadsheet users to make the “JMP” to becoming JMP data ninjas. 

 

 

Hi,  I'm  Trish  Roth.  I  am  going  to  be  presenting  to  you  today  about  managing  a  learning  program  with  JMP  that  is  developing  the  next  generation  of  JMP  Ninjas.  A  little  bit  about  myself.  I  am  a  data  scientist  in  core  diagnostics.  My  colleague,   Jeff Pennoyer,  who  helped  develop  this  training  and  the  presentation  materials,  isn't  able  to  join  us.  But  acknowledge  his  contributions  and  also  many  other  colleagues'  contributions  over  the  years  in  putting  together  training  to  improve  the  skill  sets  that  we  have  within  our  organization  with  the  use  of  JMP.

We  both  have  biochemistry,  technology  backgrounds,  and  have  worked  in  the  data  science,  analytics  space  for  a  number  of  years,  along  with  many  other  folks  in  our  division.  A  little  bit  about  Abbott,  in  case  you  aren't  familiar.  It's  a  large  global  health  care  company.  We've  been  in  business  for  over  130  years,  operate  around  the  world.

We  have  over  113,000  employees.  We're  all  focused  on  bringing  life- changing  health  technologies  to  the  people  who  need  them.  Y ou  can  see  on  the  right  hand  side  a  number  of  the  different  product  lines  that  we  support.  It  varies  by  country.  If  you  go  to  abbott.c om  from  your  location,  you'll  see  more  information  about  the  kinds  of  products  that  Abbott  delivers.

Both  Jeff  and  I  and  a  number  of  colleagues  who've  been  involved  with  this  project  and  training  are  from  the  diagnostics  division,  particularly  core  laboratory  where  we  work  with  large  hospitals   and  reference  laboratories  who  provide  diagnostic  testing  directly  to  patients  or  to  physicians.  Y ou  can  see  some  of  the  other  product  lines  here.

T he  purpose  of  the  presentation  is  really  twofold.  Wanted  to  give  some  insight  and  thoughts  around  how  we  approach  training,  the  types  of  things  we  include.  How  we  organize  it,  and  as  well  talk  about  some  of  the  features  and  functions  of  JMP  that  we  focus  on,  particularly  in  our  beginner  training,  to  get  people  comfortable  with  data  manipulation  and  data  preparation  and  data  summarization.

These  skill  sets  can  really  serve  them  as  they  continue  to  grow  and  develop  as  data  analysts  and  move  on  to  more  advanced  analytics  and  statistical  analysis  and  modelling.  But  this  is  a  good  foundation  to  get  people  started  and  comfortable.  T hat's  the  approach  that  we  have  taken.  We  leverage  materials  that  JMP  provides.  We  leveraged  area  experts.  We  try  to  have  area  specific  examples  to  really  make  the  training  relevant  to  people  so  that  they  can  see  the  value  and  how  they  might  apply  it  in  their   day-to-day.

We  talk  to  both  managers  and  employees  about  what  they  want,  what  they  need.  As  we  try  to  think  about  how  much  can  we  really  deliver.  What  kinds  of  skill  sets  are  we  lacking,  or  where  do  we  not  have  enough  people.  How  do  we  grow  those  skill  sets.  How  do  we  fill  those  knowledge  gaps.  Then  on  the  employee  side,  people  want  to  contribute.  They  want  to  grow.  They  want  to  become  functioning  members  of  their  departments,  especially  when  they're  new.  They  want  to  be  independent.

H opefully  we  can  find  an  intersection  between  those  needs  and  wants  to  develop  some  training.  We  always  have  to  have  the  conversation  around  investing  time,  both  from  a  manager  standpoint  and  giving  their  employees  time  and  space  to  work  on  developing  their  skill  sets  and  employees  have  to  be  willing  to  invest  time  and  practice  and  think  about  how  they're  going  to  apply  things  so  that  it  sticks  and  they  really  do  hone  their  skill  sets.  We've  defined  a  body  of  knowledge  that  we  focus  on,  particularly,  again,  at  the  beginning  and  intermediate  stages.  We  have  a  fair  amount  of  information  and  knowledge getting  started  and  we  do  this  in  what  we  call  a  boot  camp  style.

Sometimes  it's  intensive  over  a  couple  of  days.  A  couple  of  hours  at  a  time  to  just  really  get  people  into  JMP,  get  them  familiar  with  how  to  work  with  data  in  JMP.  Oftentimes,  they're  coming  from  Excel,  so  we  have  to  reorient  them  a  bit  but  the  basics.  What  are  the  menus?  What  are  the  preferences?  How  do  you  get  data  in?  How  do  you  do  some  basic  data  clean-up  and  summarization  functions,  basic  graphing,  creating  formulas?  T his  allows  people  to  get  up  to  speed  and  be  able  to  actually  deliver  some  analysis  pretty  readily  once  they  get  through  this  boot  camp  core  information.

Then  depending  on  how  deep  we  want  to  go  with  the  learning.  Depending  on  what  the  organizational  needs  are.  What  the  time  availability  is.  We  will  start  to  get  into  the  more  traditional  exploratory  data  analysis  and  statistical  analysis  like  particularly  capability  analysis,  control  charting,  hypothesis  testing,  regression.  A s  people  move  into  these  topics,  they  then  go  on  to  do  much  more  modelling.  We've  got  a  lot  of  interest  in  scripting.  O nce  people  have  these  foundations,  they  can  start  to  move  on  to  these  other  topics  and  really  deliver  value  to  the  organization.

Once  we  got  the  material  and  how  much  we  think  we're  going  to  deliver  and  present,  we  have  to  really  think  about  what's  the  best  way  to  deliver  it.  We  do  like  the  in  person.  Obviously,  in  the  last  couple  of  years,  we've  not  done  a  lot  of  that.  But  there  are  some  folks  who  just  really  do  better  face  to  face  where  they  can  have  somebody  standing  over  their  shoulder  and  watching  what  they're  doing.  We  do  predominantly  use  virtual  conferencing.  We  can  bring  together  people  from  a  lot  of  different  sites,  locations  that  way,  and  minimize  travel.

We  also  do  some  very  informal  things,  small  bursts  of  topics,  one  particular  topic,  maybe  over  a  lunch  hour  or  a  small  group  meeting.  We  also  have  some  fully  independent  learners.  We  point  them  to  the  learning  resources  both  from  JMP  and  also  from  our  curated  set  of  information  and  presentations  and  recordings  that  we  have  internally.  W e've  planned  for  how  to  organize  and  centralize  and  post  information  so  that  it  remains  accessible  to  others  when  they  want  to  come  and  do  some  training.

Just  have  a  little  snippet.  From  our  SharePoint  side,  just  basic  information.  You  don't  have  to  be  a  really  good  website  developer.  You  can  put  a  little  calendar  of  events.  You  can  have  information  for  beginners,  intermediates.  We  provide  links  to  past  recordings.  When  folks  finish  their  training,  we  like  to  do  a  little  congratulations  and  give  them  some  recognition.  I nternally,  these  links  would  take  you  out  to  some  presentations  to  listing  folks  who  have  successfully  completed  elements  of  our  training.

But  then  this  leaves  a  body  of  knowledge  and  a  body  of  resources  internally  that  folks  can  leverage  as  well.  We  have  a  lot  of  links  out  to  the  JMP  community  where  there's  a  lot  of  good  information  and  SharePoint  document  libraries,  so  presentations  and  data  files.  We  can  keep  this  all  centralized  and  people  can  access  it  on  demand  when  they  have  time  or  interest  in  training.

A s  well,  we  maintain  a  list  of  subject- matter  experts  here  just  showing  Jeff  and  I.  But  there's  many  other  colleagues  that  have  been  involved  and  give  their  time  and  talent  to  help  others  develop.  They  put  a  direct  link  to  top  five  countdown  of  why  data  preparation  is  faster,  easier,  and  better  in  JMP.  It's  about  a  three  minute  video  from  Julian  Paris  at  JMP.  It  gets  people  energized  and  motivated  and  excited  when  they  see  all  of  the  features  and  functions  that  they're  going  to  be  learning  about. W e  sometimes  kick  off  training  with  a  couple  of  little  videos.

As  I  mentioned,  we've  defined  some learning  levels  that  helps  folks  try  to  figure  out  what  they  should  sign  up  for  or  where  they  might  fit.  It  is  a  challenge  because  there's  such  a  broad  base  of  functions  within  JMP  that  an  intermediate  level  or  beginner  level  can  cover  a  lot  of  territory.  But  we  do  our  best  to  try  to  get  people  into  a  group  where  they  feel  comfortable  and  are  at  the  same  learning  pace  and  level.  O ur  more  advanced  and  intermediate  folks,  we  have  them  do  teach  facts  and  presentations.  It  helps  hone  their  skills  and,  again,  builds  the  community  within  our  organization.

Put  a  little  example  of  how  we  survey  to  solicit  people  that  might  be  interested  in  the  training.  We  leverage  Microsoft  Forms.  We  can  create  internal  surveys,  collect  demographic  information  about  people  who  their  managers  are.  Again,  it's  very  important  that  there's  good  collaboration  and  communication  between  the  learners  and  their  managers  to  make  sure  that  this  is  something  that  can  be  supported.  We  need  to  know  where  people  are  so  we  can  consider  time  zones  as  we're  thinking  about  how  we're  going  to  schedule  training.

Just  survey  101,  the  more  you  can  give  canned  responses  that  a  user  selects  from  versus  entering  their  own  information,  the  easier  you  will  have  in  being  able  to  analyze  and  summarize  that  information  when  you  get  it  back.  Have  a  multi  response  question,  because  one  of  the  things  we  can  look  at  it  shortly  when  we  get  to  the  demo  is  how  JMP  can  handle  a  single  question  that  has  multiple  responses  so  that  you  can  understand  different  categories  that  people  might  have  selected.

But  we  do  also  want  to  understand,  again,  the  why  folks  want  to  participate  in  the  training  so  we  can  ensure  that  we  meet  their  needs  and  that  they're  coming  into  it  for  the  right  reasons.  S ometimes  we  collect  other  information  about  other  things  they  might  be  interested  in  learning.  Now  we'll  move  into  JMP.  W here  we're  going  to  start  is  we've  done  a  survey  and  we've  gotten  back  our  results.  We're  going  to  get  that  survey  information  into  JMP  to  find  out  who's  interested  in  taking  the  class.  Then  we're  going  work  up  through  a  series  of  columns  and  formulas.  How  are  we  going  to  keep  track  of  these  folks  as  they  move  through  their  training?

I  will  be  using  JMP  17  standard,  but  most  of  this  has  also  been  done  actually  started  out  in  15  and  16.  It  will  work  there  as  well.  Just  so  you  can  see  where  we're  going.  Again,  we're  going  to  import  this  information.  We're  going  to  do  some  cleanup.  We're  going  to  enrich  the  information  by  adding  some  formulas  and  columns.  Then  we'll  create  a  subset  so  we  can  track  our  beginners.  Then  we're  going  to  start  to  import  information  into  that  table  to  keep  track  of  what  people  have  completed.  Then  I've  got  some  scoring  formulas  so  I  can  figure  out  who  has  completed  the  training.  If  not,  what  elements  of  the  training  that  they're  missing.  Then  we  can  use  that  data  table  with  the  scoring  to  then  communicate  back  congratulations  to  both  the  student  and  their  manager.

I'm  going  to  get  out  of  PowerPoint  and  go  to  a  JMP  journal.  For  the  remainder  of  the  discussion,  we  will  be  in  JMP.  Again,  the  registration  form  comes  back  in  the  form  of  an  Excel  file.  It's  embedded  in  this  worksheet.  When  you  launch  it,  JMP  is  going  to  look  to  import  that  information.  As  a  best  practice,  I  always  click  on  the  Restore  Default  Settings.  This  is  a  fairly  simple  worksheet.  It  only  has  one  tab  and  you  can  quickly  evaluate,  are  the  columns  looking  right?  My  headers  in  the  right  place?  The  data  elements  look  like  they're  going  to  be  properly  imported.  W e  have  a  quick  look  at  the  data.

If  we  did  have  any  hidden  rows  or  columns  or  empty  rows  columns,  we  could  decide  whether  or  not  we  wanted  those  to  be  imported  or  not.  We're  going  to  leave  the  defaults  for  this.  Simply  click  Import  and  there  we  go.   JMP has  ingested  the  information  from  the  spreadsheet.  Again,  there  is  83  respondents.  You  can  see  all  of  the  categories.  These  were  the  questions  in  the  questionnaire.  Each  one  comes  in  as  a  different  column  in  JMP.

Obviously,  this  is  anonymized,  so  you  can  see  the  learner's  email,  first  name,  last  name.  This  email  is  going  to  be  important  because  that's  going  to  be  the  key.  That's  going  to  be  the  piece  of  information  when  we  look  at  importing.  Did  they  attend  a  training  session?  Did  they  turn  in  their  homework?  That's  going  to  be  how  we  join  the  information.  All  of  the  reports  that  we  get  will  have  the  learner's  email.  That's  how  we  can  combine  data  that's  going  to  come  as  we  progress  through  the  training.

What  are  we  going  to  do  with  this  file?  We've  imported  it.  We're  going  to  do  some  data  functions  to  clean  it  up  and  enrich  it.  I've  listed  those  here.  We're  going  to  look  at  the  location  information  and  we're  going  to  see  that  we  have  some  permutations.  We're  going  to  use  the  Recode  function  to  clean  that  up.  When  we  surveyed,  we  combined  track  and  level.  When  we  did  this  survey,  we  were  actually  surveyed  for  more  than  just  JMP  training.  This  is  a  subset  of  that.  But  we  want  to  separate  those  into  two  pieces  of  information  rather  than  having  them  glued  together.

Then  we'll  take  a  look  at  doing  some  summaries  and  tabulations  and  graphics  so  that  we  understand  who  is  the  learner  population  that  has  signed  up  for  training.  I'm  going  to  jump  over,  save  this  one  to  the  version  where  I've  already  cleaned  this  up.  W e'll  take  a  look  at  what  that  looks  like.  Again,  here is,  in  the  survey,  we  have  location  and  you  can  see  there's  a  new  column  with  a  plus  sign  that's  got  a  formula. Y ou  can  have  a  look  at  it  and  see  that  it's  doing  some  manipulation.  Where  it  said  Lake  Forest,  we're  actually  converting  that  to  Chicago,  Illinois.

The  Recode  function  generated  those  formulas.  I  did  not  have  to  write  that  formula.  The  way  that  you  do  that  is  you  go  to  learner  location  and  you  can  just  right- click  and  select  Recode.  It's  going  to  show  you  here's  all  the  data  elements.  You  can  see  pretty  quickly  that  there's  some  permutations.  Somebody  entered  Knoxville,  Tennessee  with  and  without  a  parenthesis.  Geographically,  you  may  or  may  not  know.  Chicago  is  a  big  location,  a  big  city,  and  actually  Des  Plaines  and  Waukegan  are  actually  all  suburbs  and  Lake  Forest  as  well  are  all  really  part  of  Chicagoland.

W e  want  to  group  those  together.  They're  all  the  same  time  zone.  Those  folks  are  within  15,  20- 30  minutes  of  each  other.  W e  could,  again,  we've  got  them  all  highlighted.  I  highlighted  multiple  by  holding  down  the  CTRL  key.  I'm  going  to  right- click  and  I  can  say  I  want  to  group  these  all  to  be  Chicago.  N ow  you  see  all  four  of  these  entries  are  going  to  be  Chicago.  I'm  actually  going  to  add  Illinois.  W e  go  through  that  process  for  a  number  of  the  different  permutations.  Again,  Santa  Clara,  California  got  entered  with  and  without  the  state  designation,  so  I  can  group  these.   I  just  want  to  use  the  two- letter  designation.

Similarly,  so  we  can  go  through  the  different  permutations  and  do  this  data  clean  up it.  The  way  that  we  want  to  save  it  is  we  could  overwrite  the  data.  I  don't  like  to  overwrite  data  in  my  data  tables.  I  want  to  save  the  formula  because  if  I  run  another  class.  Or  I  go  to  the  run  another  survey,  it's  likely  that  I  might  see  similar  permutations.  Just  for  the  purposes  of  demo,  I'm  just  going  to  rename  this  as  demo.  Now  it's  going  to  create  a  new  column  formula.  I'm  going  to  hit  Recode  and  you  can  see  that  it  created  this  new  column  here  with  the  formula.

A gain,  I  didn't  do  all  of  the  clean-up  permutations,  but  you  can  see  how  it  did  the  mapping.  I n  the  future,  if  it  sees   Des Plaines,  it's  going  to  group  all  of  these  under  Chicago.  W hy  do  that?  It  obviously  reduces  the  number  of  variables  if  you  try  to  plot  or  summarize,  and  it  just  cleans  things  up. W e  did  some  additional  clean  up  items.  As  I  mentioned,  this  track  and  level,  it  broke  it  into  two  pieces  using  a  word  formula.

We  just  said  take  the  first  word  of  track  and  level,  and  there  you  have  it.   JMP and  take  the  last  word  of  track  and  level,  and  that  will  get  you  level.  You  can  do  that  by  simply  taking  the  combined  data  column  and  using  some  pre-set  column  formulas  that  JMP  provides.  Here's  first  word,  or  you  can  select  last  word,  you  can  see  first  word  JMP.  Then  I  just  retitled  these  to  simplify  the  name.

Now,  why  do  I  do  that?  Now,  I  want  to  have  a  look  at  what's  in  this  data  set  so  I  can  use  the  Analyze,  Tabulate  function.  Now  that  I  have  them  separated,  I  could  leave  them  like  this.  Y ou  can  see  the  population  of  beginners  and  intermediates.  I  like  to  have  them  broken  out.  I'm  going  to  do  track  and  then  I  want  to  know  level.  T here's  different  drop  zones  where  you  can  put  these  depending  on  what  you  want  to  see.  But  I'm  going  to  build  up  a  series  of  charts.  I'm  going  to  do  location.

Now  you  can  see  that  there  were  83  respondents,  61  beginners,  22  intermediates.  If  I  check  the  box  down  here  for  order  by  by  count  of  grouping  columns.  You  can  see  that  it  resorted  so  that  the  grouping  with  the  highest  count  is  listed  first.  So  rather  than  being  alphabetical,  it's  in  descending  order  by  how  many  are  in  each  of  the  categories.  You  can  quickly  see  what  is  the  distribution  of  the  locations  of  the  people  interested  in  your  training,  and  you  can  start  to  plan  for  how  you're  going  to  deliver  it.

If  you're  done,  you  can  click  done.  Then  what  I've  done  is  use  the  script  function  within  JMP,  Save  script  to  data  table.  Then  I  can  give  that  a  descriptive  name  and  it  will  save  it  right  back  to  the  data  table.  Here  we  go.  Tabulate  by  location  and  level.  If  I  click  that  button,  basically  it's  repeating  that  analysis.  I've  added  a  little  bit  more  detail  where  I,  in  addition  to  the  number  of  respondents,  percentage  of  total,  so  you  can  see  what  proportion  is  in  each  category.

Instead  of  tables,  graphics  are  always  nice.  Again,  I've  presaved  some.  We'll  take  a  look  quickly  at  how  to  build  those.  I've  taken  advantage  of  a  function  called  the  column  switcher.  Built  up  a  graph  that  I  liked  and  now  I  can  easily  toggle  between  different  categories.  You  see,  this  one's  a  little  bit  messy,  but  you  can  go  between  categories  to  have  a  look  at  different  managers.

Some  only  submitted  one  person  to  be  going  to  training.  Some  submitted  multiple.  It's  a  little  bit  busy.  But  again,  you  can  see  what  functional  area  they  come  from  and  you  can  see  what  location  they  come  from.  I've  also  added  a  data  filter.  If  I  really  just  wanted  to  hone  in  on  beginners,  I  could  select  beginner  intermediate.

You  can  see  in  one  page,  I  can  very  quickly  get  a  variety  of  graphs  and  easily  put  them  into  a  presentation  or  save  them  so  I  can  communicate.  The  beginners  are  mostly  from  Chicago,  but  there's  a  good  chunk  from  Dallas.  We've  got  two  Irish  sites  that  have  a  number  of  folks  that  are  interested.

But  if  I  go  to  intermediate,  Chicago,  Germany,  again,  Texas,  you  can  see  where  folks  are  coming  from  that  are  interested  in  different  levels  of  training.  The  way  we  build  this,  I'll  pull  it  off  to  the  side  and  we'll  just  look  at  it  quickly.

Graph  Builder.  I  started  with  location.  I'm  actually  going  to  drag  location  to  the  Y  axis  and  hit  a  bar  chart.  Now  I  see  each  of  the  categories.  The  reason  I  like  to  do  this,  sometimes  if  the  text  is  long,  it's  a  little  bit  easier  to  read  it  when  it's  in  this  horizontal  orientation  than  if  it's  in  a  vertical  orientation  and  you're  trying  to  read  it  sideways.

Again,  if  you  right  click,  I  can  change  the  ordering  and  then  go  order  ascending.  A gain  now  it's  count  data.  So  it's  putting  the  category  with  the  highest  counts  at  the  top  and  the  least  counts  are  at  the  bottom.  I'm  going  to  change  this  from  mean,  it's  not  really  a  mean  it  is  just  a  count.  Now  I  can  add  a  label  that  is  percent  of  total  values.

I  can  see  again,  Chicago  is  40  %  and  you  can  get  the  proportion.  Then  on  the  Y,  X  axis,  rather,  you  can  see  the  actual  counts.  You're  getting  a  lot  of  information  in  one  graphic.  One  other  feature  that's  really  nice  in  Graph  Builder  is  I  right  click,  add  caption  box,  and  I  right  click  it  again  and  change  the  caption  box  location,  the  Y  position  we'll  put  it  at  the  bottom.  Just  there's  more  real  estate  at  the  bottom.

Again,  you  can  see  there  were  83  participants,  40 %  are  from  Chicago.  A  lot  of  information  in  one  graphic.  You  can  further  customize  it  by  right  click,  hide.  I  don't  really  want  all  these  annotations.  I  think  it's  obvious  that  it's  count.  I'm  just  going  to  hide  some  of  the  annotations.  It  just  makes  it  a  little  cleaner.

The  way  I  got  the  coloring  was  to  drag  location  over  to  color.  Now  each  category  has  got  a  different  color,  but  I  can  customize  those  so  I  can  click  on  the  bar.  Chicago's  blue,  it's  fine.  Next  one  down  is  Longford.  If  I  right  click,  I  can  change  the  coloring.  It  will  go  a  little  bit  lighter  blue.  Dallas,  Texas  was  the  next  category,  fill  color.  I  will  do  light  blue.

Then  the  rest  of  them,  I'm  just  going  to  hold  down  my  CTRL  key  as  I'm  clicking  through  all  of  these  categories  to  highlight  the  rest  of  them.  Again,  pick  one  of  them  in  the  legend,  right  click,  and  we're  just  going  to  send  them  to  gray.  That's  how  I  got  to  the  customization.  Now,  I  don't  really  need  to  see  this  legend,  so  I  can  get  rid  of  it.

The  way  the  column  switcher  works,  if  I  look  up  here  in  the  toolbar  column  switcher,  it's  also  icon.  It's  right  now  on  location.  I  can  add  to  that  organization  manager's  email,  again,  holding  the  CTRL  key.  Okay.  T  hat's  what  brings  up  this  window  where  now  I  can  switch  between.  It  gets  a  little  busy  and  I  have  to  go  through  the  same  color  customization  if  I  want  to  have  the  blue  and  just  a  simple  blue  and  Gray.

But  it  remembered  it  for  location.  I  can  say  done.  Again,  that  gets  me  real  estate.  I  don't  really  want  to  look  at  that  legend.  U nder  Show,  I  can  turn  it  off  and  then  resize.  If  I  wanted  to  be  able  to  toggle  again  between  beginner  and  intermediate,  we  can  use  a  local  data  filter  and  say  I  want  to  be  able  to  filter  on  level  to  select  it,  do  the  plus  sign.

One  nice  feature  I  think  I've  noticed  in  17  is  if  I  resize  this,  it  enlarges  the  font  automatically,  which  is  actually  very  nice.  Now  I  can  toggle  between  beginner  and  intermediate,  or  I  can  clear  the  selection  and  again  leave  it  at  both.  Once  I'm  happy  with  that,  again,  onto  the  red  hotspot,  Save  script  to  data  table,  give  it  a  descriptive  name  so  you  know  what  it  is,  and  it  will  save  that  script  to  the  data  table  so  you  can  re  execute  it.

The  other  thing  you  can  do  is  under  the  Edit  menu  is  Edit  Journal  or  CTRL  J,  then  it  will  grab  an  image  of  that  analysis  and  place  it  in  your  journal,  which  is  what  I  have  done  here. I f  I  go  to  the  journal,  you  can  see  that  I've  captured  these  images  of  the  graphs  the  way  that  I  like  them.  Then  nice  thing  about  doing  it  in  a  journal  versus  grabbing  a  static  picture  is,  again,  you've  got  your  in  JMP  and  you've  got  your  red  hotspot,  which  means  you  can  have  interactivity.

I  can  select  the  graph  from  the  journal,  and  as  long  as  the  table  behind  the  graph  is  open,  you  can  say  run  a  new  window  and  I  get  back  my  interactive  graph.  If  I  wanted  to  make  some  additional  changes,  change  the  text,  I  could  do  that.  But  I've  got  everything  saved  in  a  nice  workbook.

Now  we  have  a  learner  list.  We  know  where  they're  from.  I  wanted  to  quickly  touch  on  the  skill  set  piece  in  this  multi  response.  The  way  I  handled  that  is  I  actually  created  a  copy  of  the  column  called  skill  set.  If  you  look,  it's  a  little  bit  hard  to  see,  but  if  you  look  carefully  at  the  original  column,  each  of  the  selected  items  was  separated  by  a  semi  colon.

JMP  can  handle  a  semi  colon  as  a  delimiter.  I  found  that  it  didn't  work  very  well  in  this  analysis.  As  a  workaround,  I  created  a  copy  of  the  column,  and  then  I  did  CTRL  F,  and  you  can  do  a  simple  find  and  replace.  I  replaced  the  semi  colon  with  a  comma,  and  JMP  liked  the  comma  a  whole  lot  better.

Now  why  do  that?  Again,  I'm  going  to  go  to  the  graph  builder.  Then  the  final  thing  I  did  was  told  JMP,  Hey,  this  column  is  actually  a  multi  response  column  rather  than  being  a  number  or  character.  It's  multi  response.  That's  what  prompts  JMP  to  look  for  that  delimiter  and  understand  that  there's  different  categories  in  that  column.

Again,  we'll  just  quickly  go to  the  graph  builder  and  you  can  look  at  the  difference.  Now,  if  I  take  multi,  you  can  see  each  category  is  only  represented  once.  Of  all  the  different  reasons  why  people  are  interested  in  participating  in  training,  each  category  gets  counted  independently  and  you  don't  see  all  the  permutations.

A gain,  right  click,  order  by.  You  can  see  what  the  most  popular  ones  people  want  to  improve  their  skill  set  so  that  they  can  be  more  efficient.  We've  been  talking  a  lot  about  storytelling  with  data,  how  to  getting  a  message  across,  how  to  drive  action  with  data  stories.  These  are  all  the  reasons  that  people  want  to  participate  in training.  Yeah,  that  multi  response  function  is  nice,  particularly  if  you're  doing  surveys.

Final  thing  we're  going  to  do  on  this  data  table  is  we're  just  going  to  take  a  subset  because  we're  going  to  focus  on  the  beginners.  A gain,  you  can  use  the  data  filter  level.  I  only  want  the  beginners.  Again,  out  of  the  83,  it's  highlighting  the  61.  I'm  going  to  select  a  set  of  columns  in  my  data.  I  don't  want  all  of  them. I don't ant that one.

Then  we're  going  to  create  a  subset  table,  subset,  and  we're  going  to  tell  JMP,  I  only  want  to  use  this  selected  columns.  It  gives  you  a  nice  preview.  Here's  the  email  address,  where  they're  from,  and  their  level,  and  we  can  say  okay.  Now  we  have  just  a  list  of  the  beginners,  where  they're  from,  basic  information,  and  this  is  what  we  can  use  to  start  to  build  a  tracker  to  say,  Okay,  these  are  the  folks  that  are  beginners.  I  got  to  make  sure  that  they  complete  their  requirements.

How  am  I  going  to  do  that?  I'm  going  to  take  you  to  the  version  where  I  already  have  this  set  up.  Let's  close  the  registration  information.  Now  we're  into  the  completing.  What  I've  done  is  taken  this  basic  data  table  that  had  the  information  about  who  registered  for  training,  and  I  started  adding  a  whole  bunch  of  columns.

Beginner  training  consists  of  five  different  classes  that  they  need  to  attend,  five  different  sessions,  three  homework  assignments.  We  assign  them  a  couple  of  STIPs  modules.  That's  Cisco  Thinking  for  industrial  problem  solving.  They're  free  courses  and  modules  available  through  the  JMP  learning  community.  We  assign  a  couple  of  them  to  the  beginners.  They  can  take  more  if  they  want  to.  I've  got  them  all  listed  here.

I  will  talk  about  these  hash  marks  in  a  minute  and  why  these  columns  aren't  blank,  like  the  homework  columns.  We  request  that  they  provide  a  data  example.  These  are  all  the  elements  of  the  training.  What  I've  leveraged  is  a  couple  of  column  features.  So  if  I  go  to  class  1  and  I  do  column  information,  what  you  can  see  is  1,  I've  used  this  list  check  function.  I've  told  JMP,  these  are  the  only  values  that  can  go  in  that  column.

That  just  helps  keep  the  data  sheet  clean.  If  I  do  any  data  entry,  it  forces  consistency  across  the  data  table.  The  other  really  nice  feature  is  called  value  colors.  Then  I've  assigned  a  specific  color  to  each  value.  Yes,  if  they  attended  class,  they  get  a  nice  dark  green.  If  our  students  were  recording,  they  maybe  they  watched  the  recording  later,  they  didn't  come  to  class  live.  Sometimes  people  tell  me  they're  out  of  the  office,  just  color  coded  that  red.

Then  the  key  feature  is  if  you  click  on  the  little  box  at  the  top  and  hit  Apply,  it  will  color  code  your  cells  similar  to  what  people  are  used  to  seeing  in  Excel,  it'll  color  code  your  cells  based  on  the  content  of  that  cell.  It  makes  it  very  easy  to  look  across  this  data  sheet  or  data  table  to  say  where  are  we  at,  how  many  people  are  missing  things,  how  many  people  are  green,  how  many  people  are  red.

Once  you  have  one  column  set  up,  you  can  use  copy  column  properties,  and  I  can  broadcast  that  across  the  remaining  four  columns  for  the  different  classes,  which  is  what  I've  done.  When  you're  using  these  value  colors,  it  puts  a  little  black  X  mark  as  an  attention  activator  to  let  you  know  that  it's  going  to  color  code  depending  on  what  you  enter  there  versus  in  the  homework  field,  I  hadn't  yet  activated  that.

These  are  all  the  elements  that  are  required  of  training.  Now  I  have  my  workbook  for  managing.  One  other  feature  we'll  talk  about  is  joining.  We  held  our  first  class.  I'm  going  to  clear  this  a  little  bit.  I  actually  already  have  this  Excel  file  open.  Microsoft  Teams  provided  me  with  a  summary  of  the  meeting.  I  had  42  participants.  Here's  who  they  are.  Here's  their  email  address.

It  does  actually  tell  me  how  long  they  were  in  the  meeting.  If  I  scroll  down  to  the  bottom,  I  can  decide  if  Learner  26  who  was  there  for  12  seconds if  they  were  going  to  get  credit  for  attending  or  not.  But  really  all  I  need  out  of  this  worksheet  is  just  their  email  address  because  that's  how  I  know  who  they  are  in  my  tracking  sheet.

I've  highlighted  it  and  I'm  just  going  to  highlight  it  in  Excel.  One  of  the  other  things  you  can  notice  is,  Learner  79  must  have  had  a  little  trouble.  They  were  in  for  one  minute  and  then  must  have  gotten  dropped  or  had  to go  back,  come  back  in.  There's  actually  two  entries  for  Learner  76  and  Learner  79.  You  always  want  to  look  at  your  data  first.

But  with  JMP  and  Join,  we  don't  have  to  worry  about  that  JMP  will  do  a  good  job  of  merging  the  information.  Another  really  fun  feature  is  the  JMP  add  in.  I've  highlighted  what  I  want.  There's  a  JMP  add  in  within  Excel,  select  it,  data  table.  It's  opening  it  on  my  other  screen.  I'll  pull  it  over.  I've  literally  just  grabbed  that  information  and  put  it  in  a  JMP  data  table.

I'm  going  to  quickly  add  a  column  called  Class  1.  This  is  not  a  data  table  we're  going  to  keep,  so  I'm  not  going  to  spend  its  character.  Okay.  Then  this  is  the  list  of  names  of  people  who  attended  class  1.  I'm  going  to  just  enter  Y  because  I  know  that  that's...  I'm  going  to  fill  to  the  end  of  the  data  table  and  it's  already  called  attendance.

I  don't  need  the  Excel  spreadsheet  anymore.  I'm  going  to  go  back  to  my  tracker.  You  can  see  that  I've  deleted  some  of  the  entries  here.  I'm  going  to  use  a  table  update  function.  I  like  update  because  I  just  keep  building  onto  the  same  table.  I  don't  constantly  generate  new  tables  that  I  have  to  rename  and  save.

In  that  attendance  list,  I  know  that  email  matches  email.  It  is  case  sensitive,  so  I  actually  had  to  make  sure  that  it  was  in  full  lowercase  in  both  locations  in  order  for  it  to  match  up.  JMP  will  give  you  a  really  good  preview.  If  you  don't  see  anything,  then  you  can  take  a  look  at  whether  or  not  maybe  you  missed  something.

Then  what  I  want  to  do  is  the  attendance  table,  which  is  the  update  table,  I  want  to  update  the  class  1  information  there.  I  want  to  replace  the  class  1  information  in  the  Master  table  because  it's  just  blank.  Let  me  see  if  I  can  do  this  so  you  can  see  what  happens  when  I  hit...  Here  we  go.  One,  it's  giving  you  a  preview,  but  if  you  watch  up  here,  this  is  the  tracker  sheet.  I'm  going  to  update  it.  I  hit  okay,  and  there  we  go.

Now  it's  updated  the  tracker  sheet  with  the  information  about,  yes,  these  people  attended  class  1.  As  we're  moving  through  a  training,  we're  going  to  do  that  on  a  repetitive  basis.  We're  going  to  get  reports  about  attendance.  We're  going  to  get  reports  about  who  completed  their  homework.

I  don't  have  to  manually  go  in  here  and,  okay,  you  were  there,  you  were  there.  I  can  do  it  in  a  much  more  automated  fashion.  Then  that  way,  if  somebody  emails  me  or  lets  me  know,  "Hey,  I  wasn't  there,  but  I  watched  the  recording,"  now  I  can  just  enter  that  manually  and  it  really  reduces  the  amount  of  manual  intervention.   Again,  a  lot  like  what  people  are  used  to  working  with  spreadsheets,  but  I  think  once  you  get  used  to  it,  you  can  actually  do  a  lot  more  here.

File,  close  this  one.  This  is  just  a  transient.  I  don't  need  to  keep  it  for  any  reason,  so  I'm  not  going  to  save  it.  Then  we're  going  to  go  to...  Now  time  has  gone  by.  We've  run  a  whole  bunch  of  classes.  You  can  see  people  who've  come  to  class.  People  have  missed  stuff.  Some  people  came  and  I  guess  decided  it  wasn't  for  them.  I've done  some  SIPs  modules.  This  is  what  it  looks  like  in  the  end.  This  is  the  accounting  of  what  all  of  the  participants  have  completed.  One  thing  that  would  be  nice  is  SIPs,  they  send  me  a  copy  of  their  certificate,  so  I  do  have  to  manually  enter  that  information.  It  would  be  great  to  be  able  to  get  a  report  that  I  could  then  just  join  in  or  do  some  easier  way  of  tracking  that,  but  to  be  determined.

Then  the  last  piece  is,  okay,  great,  I  know  who  was  there.  I  know  what  they  did.  Now  I  got  to  score  it.  This  is  where...  Again,  we've  got  the  tracking  spreadsheet  and  we  come  all  the  way  over  to  the  end.  Now,  I've  added  a  whole  bunch  of  other  columns  which  are  based  on  formulas  and  added  color  coding.  It's  relatively  a  little  busy,  but  it's  relatively  easy  to  see  how  much  green  there  is  versus  red.  I've  got  things  color  coded  so  it  highlights  to  me  who's  missing  information.

T hen  I  actually  even  created  a  formula  for  each  person.  What  exactly  is  missing?  If  nothing's  missing,  it's  just  a  series  of  commas.  And  then again,  a  conditional  formula,  and  we'll  look  at  these  really  in  the  moment  that  tells  me,  hey,  did  they  meet  the  minimum  requirements?  All  the  things  that  I  said  that  they  had  to  do.  If  they  did,  then  I  get  a  finished,  and  so  it's  really  easy  for  me  to  then  say,  okay,  I  know  who's  finished,  I  know  who  hasn't,  and  it  updates  automatically.  A ll  this  information  will  be  available  next  time  I  run  the  training.   Once  I  built  it  once,  I  can  make  minor  modifications,  and  so  it  becomes  a  really  helpful  tool.

Just  to  finish  out,  the  power  of  the  formula  building.  Here's  class  score.  I  said  we  held  five  classes.  I  decided  you  had  to  at  least  make  it  to  a  minimum  of  three.  How  did  I  score  that?  I  created  a  column  called  class  score.  Again,  you  can  see  it's  got  a  formula.  We'll  take  a  look  at  the  formula.  It  looks  pretty  busy,  but  we  can  build  it  up  in  pieces  and  show  you.  Once  you  get  accustomed  to  building  logic,  you  can  copy  and  paste  the  elements  and  replicate  them  pretty  quickly.

I f  we  just  take  a  really  quick  look,  each  box  is  a  different  class.  They  come  to  class  one.  If  there's  no  entry  in  that  field,  it  means  they  didn't  come.  If  the  entry  is  not  a  Y  and  it's  not  an  R  because  remember,  Y  stands  for  yes,  R  stands  for  recorded.  I f  they  didn't  come  to  class  directly  or  participate  in  the  recording,  they  get  zero  points  for  class  one.  If  they  had  a  Y  or  an  R  in  these  different  ways,  you  could  write  the  logic,  they  get  a  point  and  you  basically  take  this  element  and  you  can  paste  it  and  then  update  same  formula  class  two,  class  three.  You  see  I'm  adding  them  up.   They  get  one  point  for  class  one,  one  for  class  two,  three,  four,  and five.

It  totals  up,  and  it  is  a  little  bit  hard  to  see  with  some  of  the  color  coding,  but  this  person  only  came  to  one  class.  This  one  came  to  all  four.  Again,  this  is  where  we're  using  the  value  colors.  They're  not  showing  an  order.  If  it's  zero,  one  or  two,  because  that's  not  the  minimum  requirement,  I  color- coded  it  in  some  red  coloring  and  three,  four  and  five,  which  is  minimum  or  above  is  green.   Again,  a  lot  of  information  you  can  build  up  in  formulas  pretty  quickly.

Same  thing  for  homework.  If  we  look  at  the...  There  it's  a  minimum  of  two.  We  can  take  a  really  use  a  slightly  different  logic  this  time  just  to  show  you   the  flexibility.  In  the  homework  field,  if  nothing's  there,  that  means  they  didn't  get  a  check mark  or  a  tick  saying  they  completed  the  homework.   If  homework  is  missing,  and  then  this  little  exclamation  point  means  not.   If  it's  not  missing,  which  means  it's  there,  they  get  a  point.  A gain,  you  sum  them  up.  So  if  they  did  all  three  homeworks,  they  would  get  three  points.

You  just  work  your  way  across  tips.  We  required  two.  Some  people  overachieved  and  you  see  this  person  at  the  top,  it's  really  dark.  Did  all  seven.   I  developed  an  extra  credit  formula  saying,  okay,  if  you  were  assigned  to,  if  you  did  more,  I'll  give  you  some  bonus  points.  And  that  way,  if  you  missed  a  homework,  you  you  can  cover  it  with  an  extra  steps  module.  A gain,  you  can  just  build  up  logic  statements.  You  have  to  really  think  through  what  your  requirements  are  and  what  the  logic  form  is  going  to  be.

We'll  just  do  a  quick  how  do  we  build  that?  I  think  I  had  this  column  seven,  which  was  an  example.  Yeah.  All  right.   We'll  just  clear  this  out,  edit  formula.  Again,  I'm  just  going  to  clean  this  up,  build  it  from  scratch  really  quick.   Again,  once  you're  in  the  formula  editor,  if  you're  not  sure  where  things  are,  you  can  type  and  it'll  show  you  if  it's  under  conditional.  Just  going  to  clean  that  up.  Then  it  guides  you,  well,  what  does  it  need?  If  what?  Well,  if...  Make  sure  you  highlight  the  box.  Class  one,  I  want  to  do  a  comparison.  Is  missing  zero  else  1.  We're  just  going  to  do  a  simpler  formula.  There  you  have  it.

Now  you  can  add.  Now  I  need  to  do  the  same  thing  for  class  two.  Depending  where  you  click,  you'll  highlight  different  parts  of  the  formula.  You  want  to  make  sure  you  get  the  whole  box.  You  can  use  your  up  arrow.  Once  this  whole  formula  box  is  highlighted,  I  can  say  Control +  C  for  copy.  Now   I  can  just  paste.  I  don't  have  to  build  up  this  if  then  else  logic.  I nstead  I  can  just  say,  okay,  class  one,  I  want  to  apply  the  same  thing  for  class  two.   That's  how  you  iteratively  would  build  up  a  formula  once  you  apply––  You  can  start  to  see  that's  how  we  added  up  the  scores  based  on  the  yeses  or  our  content  being  in  the  attending  class  formula.

Then  the  final  piece  is  I  want  to  know  what's  missing.  One  thing  you'll  notice  is  that  this particular  row  here,  notice  that  they  aren't  designated  as  having  finished  their  training.   If  I  look  across  the  row,  they  completed  four  classes,  they  completed  all  three  homeworks,  they  actually  completed  four  SIPs  modules.  I  know  it's  difficult  to  see  with  the  coloring,  but  what  they  did n't  complete  was  providing  a  data  example.  It's  blank.

That  is  a  mandatory  element  of  completing  the  course.  Even  if  they  overachieved  on  everything  else,  unless  they  apply  their  learning  and  provide  us  with  some  example  of  how  they  use  JMP,  they  can't  get  full  completion  credit.   That's  why  even  though  they  have  the  points,  they're  missing  the  one  critical  element.

These  formulas  are  stored  in  the  table  that  you  can  reference  later,  but  it  gives  you  an  idea  of...  You  can  build  up  some  pretty  complex  formulas,  but  it's  saying,  okay,  if  their  class  score  is  less  than  three,  that  means  they  didn't  attend  enough  classes.   Note  that  class  is  one  of  the  items  missing.  These  double  pipes  are  for  concatenate.  It's  just  putting  a  comma  delimiter  between  the  elements.  Then  it's  saying  if  homework  is  less  than  two,  that  means  they  didn't  finish  the  minimum  number  of  homeworks  and  so  on.

Then  you  can  see  this  data  example.  If  data  example  is  missing,  it's  not  points,  it's  just  black  or  white.  It's  either  there  or  it's  not  there.  You  can  get  a  listing  of  what  they've  completed  and  what  they've  not  completed.  Once  you  have  that,  you  can  quickly  tabulate and  we'll  just  go  to  missing.  Then  I  can  add  email.  Now,  because  I  did  it  in  the  opposite  order,  here's  people  that  are  missing  one  SIPs  Module.  Here's  people  that  are  missing  two,  here's  people  that  are  missing  two  SIPs  modules  and  didn't  do  a  data  example.  Then  I  can  communicate  back  out  to  those  groups  of  folks  exactly  what  they're  missing,  and  so  they  can  either  get  it  done  or  say,   I'm  not  going  to  be  able  to  finish  this.

Close  this. Once  we've  spent  the  time  to  build  up  those  formulas,  again,  we  can  do  some  graphics  based  on  that  finished  column,  and  I  can  see  what  percentage  of  people  by  site  or  location  finished  the  training.  I  can  tabulate  35...  We  had  just  over  50 %  completion  rate.  Not  great,  but  that's reality  and  we  can  circle  back  with  what  were  your  barriers  to  finishing?  Again,  you  can  look  at  your  metrics,  you  can  report  back  on  what's  happening,  all  being  driven  off  of  this  one  data  table  by  using  different  formulas  and  different graphics. It's  very  simple  bar  charts  and  summary  tables.

Hopefully  that  gives  you  a  flavor  of  without  getting  into  advanced  analytics  and  model  building  and  response  surface  modeling,  you  can  get  a  lot  of  mileage  out  of  the  fundamental  features  of  JMP.   It's  really,  in  my  mind,  a  very  good  jumping  point  for  folks.   We've  had  a  lot  of  success  with  getting  people  up  and  running  and  comfortable.

If  you  can  navigate  through  these  tabulations  and  summaries  and  data  cleanups  and  making  some  graphs  and  customizing  the  graphs  and  thinking  about  how  to  annotate  the  graphs  so  they  have  a  quick,  meaningful  message  in  the  most  crisp  presentation,  you  will  have  really  moved  the  needle  on  the  capabilities  of  your  organization, and   hopefully  generated  some  excitement  for  the  use  of  JMP.

With  that,  I  thank  everyone  for  tuning  in.  Hopefully,  when  this  is  posted  in  the  community,  if  you  have  questions,  thoughts  or  suggestions,  certainly  welcome  the  discussion  and  hearing  what  other  people  have  to  say.  But  don't  undervalue  how  far  you  can  get  with  getting  a  broad  base  of  beginners  up  and  running.  They  can  go  out  and  do  great  things,  as  I  said,  by  way  of  summary,  get  people  excited  and  get  people  up  and  running.

The  nice  thing  is  you  get  beginners  and  advanced  practitioners  now  on  the  same  platform.  They  can  start  to  talk  to  each  other.  The  beginners  can  move  along  and  the  advanced  practitioners  don't  have  to  go  backwards.  We're  trying  to  remember  how  Excel  works,  they  can  stay  in  the  platform  where  they  do  most  of  their  analytics.  When  you  do  that,  you  can  join  the  Ninja  community  and  dare  mighty  things  like  flying  helicopters  on  Mars.  Thank  you  very  much.

Article Tags