• No results found

A few points were reiterated by many of those individuals that PERO worked with in the development of this tool:

N/A
N/A
Protected

Academic year: 2021

Share "A few points were reiterated by many of those individuals that PERO worked with in the development of this tool:"

Copied!
16
0
0

Loading.... (view fulltext now)

Full text

(1)

Practical  Framework  for  Analysing  Impacts  of  Online  Engagement  

In developing a framework for researchers to analyse their research, the Public Engagement with Research Online (PERO) team sought out feedback and suggestions from both academic

researchers and BCE practitioners by holding workshops, primarily within the University of Warwick. The case study (focusing on the CAGE research group) employed quantitative and qualitative analysis using web-based public discussion responding to research, framed within a theoretically and methodologically robust impact evaluation framework.

A few points were reiterated by many of those individuals that PERO worked with in the development of this tool:

• The tool had to be simple to understand.

• The framework needed to be explicit in how it was to be used.

• If necessary, the framework needed to identify sources of training in particular methods.

Therefore, we have proposed a framework, which is straightforward to implement, yet capable of producing robust and valid impact evaluation results. This framework illustrates the impacts of seeking to engage publics with an original piece of research online.

The framework consists of four steps:

1. (If appropriate), generate and insert Google Analytics code on relevant (e.g. institutional or personal) webpages communicating research ideas that are the subject of the impact evaluation.

2. Gather and validate keywords from Google Analytics, the web, and/or in-person events. 3. Use the acquired, validated keywords to gather online public discussions, in public spheres,

that reference key themes and/or original research, quality checking the resulting data. 4. Analyse randomly selected webpages / discussions qualitatively and/or through quantitative

content analysis.

 

Google   Analy+cal  

Code  

• Register  for  Google  Analy+cs  and  develop  code.  

• Insert  Google  Analy+cal  code  into  all  of  the  relevant  web  pages.    It  is  best  to  do  this  before  research  is  desiminated   online.    This  allows  the  researcher  to  monitor  the  visitor  traffic  and  keywords  before  and  aAer  the  dissimina+on.  

Capture   Keywords  

• Develop  a  list  of  keywords  using  the  results  from  the  Google  Analy+cs  output  

• The  keywords  should  be  of  keywords  provided  by  Google  Analy+cs  and  keyords  taken  from  the  visitor  source   pages.  

• To  obtain  the  keywords  from  the  visitor  source  webpages,  run  the  ar+cle  text  through  simple  text  analysis  tool.    

Iden+fy   Webpages  

• Using  the  keywords  conduct  a  web  search  to  iden+fy  web  discussions.       • Confirm  that  each  selected  web  discussion  is  discussing  the  relevant  case   • Record  the  iden+fidy  webpage's  web  address  in  a  database.  

Qualita+ve   Analysis  

• Randomly  select  websites  from  the  database  for  qualita+ve  analysis.    

• For  qualita+ve  methods  and  analy+cs  skills  training,  we  suggest  taking  the  ‘Evalua+ng  Impacts  of  Public   Engagement  and  Non-­‐Formal  Learning:  Qualita+ve  Methods’  workshop.  

• To  effec+vely  measure  the  impact  of  dissemina+on  online,  the  procedure  of  either  qualita+ve  or  quanita+ve   content  analysis  online  discussions  at  these  web  addresses  would  need  to  occur  prior  to  and  following   dissemina+on.  

(2)

Case  Study  

This framework of online public engagement impact evaluation has been applied to a specific instance of online public engagement conducted by Professor Andrew Oswald. Oswald is an applied economics and quantitative social scientist at the University of Warwick. His research has focused on the economic and social determinants of happiness and well-being. In 2004, Oswald released a piece of research "Money, Sex, and Happiness: An Empirical Study" with David

Blanchflower in the Journal of Economics, which was then followed up with a number of offline and

online public engagement events, including:

• Organiser, International Conference on Happiness, Adaptation, and Prediction at Harvard

University, 2007

• "Happiness PowerPoint Talk: Esmee Fairbairn Lecture, Lancaster", November 2006

• "How did we get into the crisis, and how will human happiness be affected?" TEDxWarwick,

February 2009

• 'Happiness, Lottery Winners and Your Heart', University of Warwick, July 2009

• ‘Modern Society and the Economics of Happiness’ University of Warwick Podcast,

September 2011              

(3)

Step  1:   Develop  and  Insert  Google  Analytics  Code  

The  first  step  in  the  framework  requires  the  researcher  setup  the  initial  tools  needed  for  analysis.    The   fundamental  tool  for  online  analysis  is  Google  Analytics.    Google  Analytics  is  an  analytical-­‐software  package   initially  designed  for  marketers  and  individuals  without  significant  information  technology  skills.  Therefore,   the  software  is  relatively  simple  to  begin  working  and  provides  numerous  free  learning  tutorials.    The   package  has  been  constructed  to  assist  users  in  conceptualising  a  number  of  metrics  (e.g.  number  of  visitors,   unique  visitors,  sources,  and  landing  pages),  which  illustrate  how  the  website  is  operating.    Google  Analytics   will  generate  detailed  statistics  on  each  of  its  metrics.    These  detailed  statistics  are  particularly  useful  in   building  a  foundation  for  online  public  engagement  research,  as  researchers  are  able  to  track  how  their   publics  are  engaging  with  their  websites  through  landing  pages,  keyword  usage,  and  sources.      

 

There  are  a  number  of  steps  needed  to  begin  this  process.  The  PERO  research  team  suggests  that  to  return   the  most  robust  results,  the  Google  Analytics  code  be  established  prior  to  the  public  engagement  event.  If   the  code  is  established  and  Google  Analytics  is  collecting  data  on  public  engagement  prior  to  the  actual   event  it  will  allow  for  the  researcher  to  conduct  a  higher  level  impact  evaluation  of  their  public  engagement   online.  

 

 In  this  case  study,  a  Google  Analytics  account  had  not  yet  been  established  for  Andrew  Oswald’s  research   websites.  The  first  action  required  was  the  establishment  of  a  Google  Analytics  account.  To  setup  an  account   on  Google  Analytics,  navigate  to  www.google.co.uk/analytics  and  click  on  the  button  ‘Create  an  Account’   (Figure  X).    Once  opened,  the  researcher  will  need  to  sign-­‐up  for  an  account  (Figure  X).      

                     

(4)

 

   

 

After  setting  up  the  account,  the  next  step  is  to  edit  the  profile,  providing  the  website  url  and  homepage   name  (e.g.  index.html,  home.html)(Figure  X).    In  the  case  of  establishing  a  profile  for  the  Andrew  Oswald   personal  webpage,  no  default  page  is  named  as  the  home  webpage  is  www.andrewoswald.com,  not   www.andrewoswald.com/index.html.    The  profile  names  referee  to  either  the  University  of  Warwick  CAGE   website  or  Andrew  Oswald’s  personal  webpage.      

                       

After  establishing  the  profile,  Google  Analytics  returns  a  unique  tracking  ID  and  tracking  code  (Figure  5).    The   code  is  the  basis  for  tracking  the  visitors  to  the  website.    The  statistics  returned  will  be  sub-­‐standard  without  

Figure  2:  Google  Analytics  –  creating  an  account.  

Figure  3:  Google  Analytics  edit  profile  screen.  

(5)

properly  entering  the  code  into  the  desired  webpage.    The  code  returned  from  Google  Analytics  was   inserted  into  every  webpage  relevant  to  this  study:    

• The  CAGE  website  homepage,  including  each  of  these  webpages:  

 Andrew  Oswald’s  individual  page    

 Happiness,  Adaptation,  and  Prediction  Conference  

• Andrew  Oswald’s  personal  website,  including  each  of  the  webpages:  

 CV  &  Research  History  

 Presentations  

 Academic  Publications  

 Media  &  Non-­‐Technical  Articles    

The  computer  programmers  for  these  websites  inserted  the  Google  Analytics  code  into  the  HTML   programme  code  for  each  of  the  webpages.    However,  inserting  the  code  can  simply  be  accomplished  by   copying  the  code  from  Google  Analytics  and  pasting  at  the  end  of  the  desired  webpage  HTML  code.      

    Once  the  Google  Analytics  code  is  inserted  into  each  of  the  appropriate  webpages,  the  researcher  can  move   on  to  step  two.    If,  however,  the  researcher  is  struggling  to  setup  the  Google  Analytics  code,  a  number  of   tutorials  and  videos  can  be  found  by  navigating  to  http://www.google.com/analytics/learn/index.html.    

(6)

Step  2:  Gathering  Keywords:  

The  second  step  in  the  framework  involves  the  establishment  of  keywords  to  be  used  in  the  detection  of   public  engagement  with  the  research  online.    Keywords  are  those  words  that  are  most  frequently  used  by   the  audience  to  engage  with  the  research.    In  the  case  of  Andrew  Oswald’s  research,  the  audience  used   words  such  as  ‘religion’,  ‘sex’,  ‘  happiness’,  ‘money’,  ‘poor’,  ‘wealthy’  and  ‘Andrew  Oswald’.  These  words  are   used  frequently  by  the  research  audience  to  discuss  the  research  and  can  be  used  to  reveal  further  public   engagement  occurrences,  which  may  not  be  otherwise  apparent  to  the  researcher.  

 

There  are  three  distinct  techniques  available  for  gathering  keywords  for  online  public  engagement  analysis.     The  first  two  methods  involve  using  Google  Analytics  output.    The  first  method  is  a  direct  capture  of  the   keywords  that  are  collated  by  Google  Analytics.    Google  Analytics  collates  a  list  of  keywords  that  the  

researcher’s  audience  uses  to  find  the  specific  piece  of  work.    In  this  instance,  Google  Analytics  collects  all  of   the  keywords  that  the  audience  for  Andrew  Oswald’s  research  use  in  search  engines  (such  as  Google,  Yahoo   and  MSN)  to  arrive  at  his  homepage  on  the  CAGE  website.    To  access  the  keyword  data  on  Google  Analytics   the  researcher  must  follow  the  following  steps:  

 

1. Log-­‐in  to  Google  Analytics  

2. Click  on  the  top  toolbar  tab  ‘Standard  Reporting’      

  3. Then,  click  on  the  left-­‐hand  toolbar  tab  ‘Traffic  Sources’  and  ‘Overview’  

 

 

(7)

4. Summary  source  data  will  appear,  including  keywords;  to  get  a  more  comprehensive  list  click  on  the   link  ‘View  Full  Report’.  

 

 

This  is  the  initial  set  of  keywords  the  researcher’s  audience  is  using  when  engaging  with  the  researcher.     However,  this  list  of  keywords  is  derived  from  a  similar  audience,  one  which  is  generally  in  search  of  specific   information  on  the  researcher  (i.e.  Andrew  Oswald).    The  number  of  keywords  selected  is  dependent  on  the   depth  of  evaluation  the  researcher  would  like  to  pursue.  In  this  case  a  list  of  20  keywords  was  recorded  in  a   spread  sheet  (Figures  6  and  7).    

   

         Figure  6:  List  of  source  keywords,  provided  by  Google  Analytics.  

(8)

 

   To  build  a  stronger  database  of  keywords,  a  secondary  tool  for  gathering  keywords  is  necessary.    Google   Analytics  offers  another  option  for  source  data:  website  source  referrals.    These  sources  are  sites  where  the   website  visitor  originated.    Keywords  can  be  gathered  from  these  original  webpages  by  analysing  the  text   within  these  sources.    To  gather  these  keywords,  the  follow  these  steps:  

1. Then,  click  on  the  left-­‐hand  toolbar  tab  ‘Referrals’  under  the  category  ‘Sources’    

   

 

2. A  list  of  the  top  10  referral  website  will  appear.    After  removing  each  of  the  host  websites,  a  list  of  8   websites  remained.    Each  referral  webpage  needs  to  then  be  analysed  for  keywords.    Using  simple   text  analysis  software,  several  free  options  exist  online;  further,  keywords  become  apparent  and  can   be  added  to  the  previous  list.    The  keyword  list  was  cleansed  of  common  words  such  as  articles,   pronouns,  and  conjunctions.      

 

The  rational  for  gathering  keywords  from  your  multiple  sources  is  to  guarantee  all  potential  target  audiences   are  included  within  the  evaluation.    Although  a  researcher  may  assume  that  they  are  aware  of  the  keywords   the  audience  would  use  when  engaging  with  their  research  they  cannot  be  sure  of  the  paths  of  public   engagement  that  may  occur.    For  instance,  with  Dr  Oswald’s  research  particular  keywords  can  be  identified   by  the  researcher:  Andrew,  Oswald,  happiness,  money,  economy,  and  Warwick.    These  are  all  terms  that  

         Figure  7:  The  list  of  Google  Analytics  keywords  cleansed  and  entered  into  a  spreadsheet.    

(9)

appear  abundantly  in  the  paper  presented  by  Dr  Oswald  and  in  the  number  of  events  in  which  he   participated.    However,  after  using  the  keyword  analysis,  as  described,  the  tool  also  pinpointed  that  the   public  also  used  terms  such  as:  religion  and  infidelity.    Without  using  an  appropriate  keyword  selection   process,  the  analysis  of  reach  and  significance  would  produce  results  that  do  not  generate  a  comprehensive   picture  of  the  researcher’s  online  public  engagement.    

   

(10)

Step  3:  Gather  webpages  referencing  keywords  and/or  original  research  

The  third  step  of  the  evaluation  framework  requires  the  researcher  use  the  established  keywords  to  gather   discussion  of  webpages  online  to  evaluate  the  public  engagement  with  the  research  online.    The  objective  of   gathering  webpage  discussions  is  to  collect  the  discourse  that  occurs  following  the  publics’  engagement  with   the  research.    Using  this  method,  the  discussion  can  be  analysed  for  common  themes,  cognitive  changes,   and  the  way  discussion  builds.    

 

To  begin  collecting  webpages  a  number  of  web  searches  need  to  be  run  on  the  keywords.  Using  Oswald’s   research  as  the  beginning  point,  top  keywords  were  run  through  Google  Search:  Andrew  Oswald  and   happiness.    A  Google  Search  on  these  keywords  returned  a  list  of  36,900  results,  as  this  number  was  large  a   random  selection  of  these  results  was  implemented.    Prior  to  performing  the  web  search  for  the  websites,  a   sample  size  was  established  of  100.    As  well,  the  web  search  was  limited  to  a  depth  of  5  result  pages,  per   web  search  (Figure  8).    To  randomly  select  for  webpages,  every  second  result  page  was  analysed.  

 

 

 

         Figure  8:  Google  Search  results  from  a  web  search  of  “Andrew  Oswald”  and  Happiness.  

Add  these  webpages,  including   the  discussion  to  sample   webpage  database.  

Results  Analysis:  depth  of  5  result   pages,  analysing  every  second   page  

(11)

 

       The  subsequent  steps  were  followed  to  build  a  database  of  public  engagement  online:   1. Open  each  webpage  

 

2. Confirm  the  webpage  relates  to  the  research  topic  and/or  that  there  is  public  discussion  on  the   topic.  

  3. Add  the  web  address  to  a  sample  database.  

 

(12)

 

To  build  a  comprehensive  database,  it  is  essential  to  repeat  the  web  search  using  a  variety  of  combinations   of  the  keywords  (e.g.  ‘Oswald  and  sex’,  ‘happiness  and  economics’,  ‘CAGE  and  happiness’).    Each  of  these   keyword  searches  will  return  a  variety  of  online  public  engagement  with  Andrew  Oswald’s  research.  Using   this  search  method  will  incorporate  unpredictable  aspects  of  public  engagement  that  were  not  previously   apparent.    Case  in  point,  “Money,  Sex,  and  Happiness:  An  Empirical  Study"  was  not  predicted  to  be  picked  up   by  religious  publics.    Yet,  religious  audiences  have  used  the  research  as  a  platform  for  discouraging  greed.    In   addition  to  using  each  of  the  relevant  keywords,  using  a  variety  of  search  engines  and  search  options  will   return  greater  diversity  of  public  engagement  results.    Google  Search  has  a  specified  ‘Discussion’  option  in   which  a  search  can  be  conducted  on  forums,  discussion  boards,  and  community  pages.    These  results  may   not  have  otherwise  been  apparent  in  a  web  engine  search.    Using  this  method  a  diverse  sample  of  online   public  engagement  populated  the  database.  The  websites  include  discussions  such  as:  

  • http://www.physicsforums.com/showthread.php?t=31238   • http://britpolanalysis.blogspot.co.uk/2011/09/economists-­‐disagree-­‐shock.html   • http://forums.windrivers.com/archive/index.php/t-­‐62952.html   • http://learning.blogs.nytimes.com/2011/03/15/what-­‐are-­‐some-­‐recent-­‐moments-­‐of-­‐happiness-­‐in-­‐ your-­‐life/                                  

(13)

Step  4:  Analyse  and  Extended  Evaluation    

The  final  component  of  the  framework  involves  the  statistical  analysis  of  randomly  selected  webpages   through  qualitative  and/or  quantitative  content  analysis.    Quantitative  content  analysis  methods  contribute   to  the  investigation  and  understanding  of  public  conceptualization,  whilst  using  broader  qualitative  analysis   identifies  the  public  discussions  of  unanticipated  publics,  who  engaged  with  the  research.    It  is  at  this  stage   that  evaluation  can  be  conducted  to  determine  what,  if  any,  influence  has  occurred  following  the  online   public  engagement  event.      

 

The  process  of  analysis  begins  by  random  selecting  of  individual  cases  from  the  webpage  database  for   analysis.  For  this  case  study,  Random.org  was  utilised  to  generate  a  random  selection  of  10  cases  (of  the   original  100)  (Figure  9).  These  figures  were  input  into  the  Random.org  software,  which  in  turn  produced  a  list   of  10  integers  between  1  and  100  (Figure  10).  These  integers  were  used  to  select  the  corresponding  case   number  in  the  webpage  database  (Figure  11).    After  selecting  the  cases  to  be  analysed,  the  discourse  from   each  webpage  (Figure  12)  needs  to  be  imported  into  a  text  document  to  be  cleansed  and  prepared  for   analysis  (Figure  13).                        

Figure  9:  To  generate  the  random  numbers  used  to  select  case  studies  for  analyses,  Random.org   was  utilised  to  produce  a  list  of  10  random  integers  between  1  and  100.  

Figure  10:  Random  sample  numbers,  generated  by  Random.org.    

(14)

                 

Figure  11:  Random  case  selection  of  Andrew  Oswald’s  "Money,  Sex,  and  Happiness:  An  Empirical   Study"  online  public  engagement  results.  

Figure  12:    Online  Discussion  referencing  the  dissemination  of  Dr  Oswald’s  research  

(15)

 

Either  quantitative  or  quality  analysis  can  be  conducted  on  the  prepared  data.    To  develop  the  tool  set   necessary  for  evaluating  the  online  discourse,  it  is  essential  that  the  researcher  (or  evaluation  consultant)   have  good  command  of  the  relevant  analytical  methodology.  Therefore,  under  the  recommendation  of  the   research  team,  persons  who  wish  to  use  either  methodology  should  enrol  in  an  evaluation  training  

programme,  such  as  the  ‘Evaluating  Impacts  of  Public  Engagement  and  Non-­‐formal  Learning’  hosted  by  the   University  of  Warwick.    These  courses  are  concise  and  targeted  towards  a  broad  audience,  interested  in   public  engagement  evaluation.      

 

Case  Study  Analysis.  In  an  exploration  of  Dr  Andrew  Oswald’s  impacts  online,  following  the  online   dissemination  of  his  paper,  “Money,  Sex,  and  Happiness:  An  Empirical  Study”,  a  series  of  distinct  online   discussions  emerged.    A  preliminary  analysis  of  one  of  these  cases  has  been  conducted  using  a  qualitative   analysis  approach.  This  case  focused  on  the  discussion  of  the  paper  between  9  individuals  in  May  of  2007.     These  individuals  belong  to  an  online  Sudanese  forum,  which  discusses  a  number  of  economic  concepts,   including:  living  in  excess,  currency  and  regulations.      

 

This  discussion  consisted  of  a  debate  between  the  9  individuals  on  the  role  of  money  on  an  individual’s   happiness.    A  number  of  the  participants  discussed  how  Dr  Oswald’s  theory  applied  to  a  number  of   situations  in  which  individuals  with  extreme  amounts  of  money  were  unhappy  or  lonely,  often  due  to  the   wealthy  individual’s  change  in  roles  within  society.    One  person  offered  the  following  example:    

 

“I  think  you  are  lonely  when  you  are  isolated  by  some  societal  status/role.    You  become  lonely  at  the  top,  not   to  be  mistaken  with  lonely  at  the  bottom  (for  the  pariah  and  the  renegades).  The  minute  you  are  rich.  Money   is  funny:  here  we  are  talking  about  extreme  wealth.  In  Sudan  if  you  are  wealthy  you  have  to  manage  many   people's  expectations  and  desires,  and  they  never  look  at  you  the  same  ever.”  

 

Whereas,  three  of  the  participants  indicated  their  disbelief  in  the  basic  premise  that  the  extremely  wealthy   were  less  happy  than  those  in  the  middle  class  or  even  those  in  poverty.  These  arguments  developed  from  a   sense  in  which  these  individuals  argue  not  all  those  with  wealth  are  lonely.    Additionally,  it  is  along  this   theme  that  the  individuals  argue  for  happiness  as  a  mind-­‐set  and  not  a  class  issue.    The  first  statement   speaks  out  against  Oswald’s  research  principle  of  reducing  the  idea  that  money  is  necessary  for  lifetime   happiness:  

 

“Great,  you  tried  the  life  of  the  rich  and  its  goodies  and  you  found  it  boring,  empty  and  lonely.  I  just  want  a   similar  chance  to  prove  for  myself  that  money  won't  make  me  happy.  I  promise  if  while  being  rich  I  felt   unhappy  and  lonely,  I  will  donate  my  wealth  to  the  people  of  Sudan.”  

   

Whilst,  other  arguments  support  the  concept  of  money  does  not  equate  to  happiness.    Instead  support  the   idea  that  happiness  is  a  state  of  mind.    

 

“I  really  see  this  very  clearly.    It  is  not  money  that  makes  people  miserable  or  happy.  It  is  the  people  

themselves.    Guns  in  the  hand  of  thugs  is  crime,  in  the  hands  of  law  enforcement  is  deterrent  to  crime.  Rain  in   the  Sahara  is  blessing,  in  New  York  is  misery.  So,  not  all  of  those  people  in  suburbia  live  miserable  lives.  Let's   forget  about  money  for  a  second  and  talk  about  happiness.  Many  people  live  their  life  thinking  happiness  is  a   destination.    When  I  buy  that  house,  marry  that  beautiful  woman,  have  those  smart  children,  then  I  will  be  

(16)

happy.  When  I  get  that  promotion,  get  the  corner  office,  and  a  marked  parking  stall,  then  I  will  be  happy.  Sadly   they  never  arrive.”  

 

These  are  examples  of  the  text  from  which  Oswald  is  cited.    The  participants  in  this  case  state  their  

understanding  and  attitudes  regarding  the  happiness  and  economy  theory;  however,  the  participants  did  not   offer  methods  of  improving  happiness,  whilst  also  creating  an  economically  secure  society.      

   

Figure

Figure	
  1:	
  The	
  Google	
  Analytics	
  homepage.	
  	
  Click	
  ‘Create	
  an	
  Account’	
  to	
  get	
  started.	
  
Figure	
  4:	
  PERO	
  Google	
  Analytics	
  profile	
  homepage	
  screenshot.	
  	
  
Figure	
  9:	
  To	
  generate	
  the	
  random	
  numbers	
  used	
  to	
  select	
  case	
  studies	
  for	
  analyses,	
  Random.org	
   was	
  utilised	
  to	
  produce	
  a	
  list	
  of	
  10	
  random	
  integers	
  between	
  1	
  and	
  100.	
  
Figure	
  12:	
  	
  Online	
  Discussion	
  referencing	
  the	
  dissemination	
  of	
  Dr	
  Oswald’s	
  research	
  

References

Related documents

Long term treatment with only metformin and pioglitazone and in combination with irbesartan and ramipril significantly ( P <0.001) reduced elevated serum

The aim of the study was to assess the presence of pathogenic microbes on treatment tables in one outpatient teaching clinic and determine a simple behavioral model for

It was decided that with the presence of such significant red flag signs that she should undergo advanced imaging, in this case an MRI, that revealed an underlying malignancy, which

Czasa− mi jednak obserwuje się pacjentów, którzy mimo optymalnego stosowania leczenia zachowawczego, wymagają częstego, ręcznego usuwania mas kało− wych z

If one has a control group and a group treated with methyl prednisolone, and they are killed at the same time (in this instance on the 21st day) one finds that the tumors

Standardization of herbal raw drugs include passport data of raw plant drugs, botanical authentification, microscopic & molecular examination, identification of

Twenty-five percent of our respondents listed unilateral hearing loss as an indication for BAHA im- plantation, and only 17% routinely offered this treatment to children with

E-book usage can be seen as a wicked problem with regards to this and the earlier-described characteristics because there is no consensus on how to reliably