2.4   Data Analysis

The  data  collected  from  the  Qualtrics  program  was  first  analyzed  through  descriptive  analysis.   The  results  for  each  section  were  analyzed  using  the  statistics  provided  by  Qualtrics  for  each   specific  question.  Once  the  data  was  organized  it  was  entered  into  the  SPSS  software.  The  data   was  cleaned  and  all  variables  were  renamed  to  match  the  research  codebook  and  


Once  on  SPSS,  specific  variables  were  either  recoded  or  computed  to  provide  specific  scores.     Descriptive  statistics  using  frequencies  with  skewness  and  kurtosis  and  cross-­‐tabulations  with   correlations  were  conducted  multiple  times  for  the  following  categories:  diagnosis  of  a  gluten-­‐ free  diet,  knowledge,  perception,  sources  of  information,  purchasing  and  dieting  behaviors  and   adherence.    


• To  identify  individuals  who  self-­‐diagnosed  themselves  versus  those  who  are  medically   diagnosed  with  either  Celiac  Disease,  Gluten  Sensitivity  or  Irritable  Bowel  Syndrome   answers  to  the  fourth  question  of  the  questionnaire,  what  best  describes  why  you  follow   a  gluten-­‐free  diet  (Q4)  were  recoded  into  two  groups:  medical  diagnosis  and  self-­‐

criteria:  Celiac  Disease,  Gluten  Sensitivity  and  Irritable  Bowel  Syndrome.  The  statistical   analyses  used  for  this  category  were  frequencies  and  cross-­‐tabulations  (Table  2).      

• Reasons  why  consumers  choose  to  follow  a  gluten-­‐free  diet  were  identified  through  the   question  why  do  you  follow  a  gluten-­‐free  diet  (Q7).  Answer  choices  in  this  question  were   categorized  into  two  groups:  medical  versus  social  reasons  and  were  analyzed  using   bivariate  correlation  analysis  (Table  3).    


• Total  Knowledge  score  was  achieved  by  creating  a  scale  for  questions  eleven  and  twelve   of  the  survey,  which  asked  the  participants  to  identify  products  and  ingredients  that   contain  gluten.  Scores  were  calculated  based  on  the  number  of  products  that  the  

respondents  got  correct.  The  score  for  these  two  questions  was  added  to  question  nine,   which  asked  the  participants  to  correctly  identify  the  definition  of  gluten.    

Question  eleven  score:  16    +   Question  twelve  score:  26    +   Question  nine  score:  1  


Total  Score:  43  (100%)    

The  total  score  was  computed  out  of  forty-­‐three  points.  Participants  who  correctly   answered  all  three  questions  obtained  a  score  of  100%.    The  mean  scores  for  total   knowledge  were  calculated  using  an  independent-­‐samples  t-­‐test  and  a  Mann-­‐Whitney  U   test  for  statistical  significance  reinforcement  (Table  4).      

• Perception  of  wheat  and  gluten  was  measured  through  a  5-­‐point  Likert  scale,  which   asked  participants  to  indicate  their  level  of  agreement  or  disagreement  with  the   following  statements:  wheat  is  good;  gluten  is  good.  The  mean  and  standard  deviation   for  these  categories  were  obtained  through  frequencies  and  cross-­‐tabulations  (Table  5).      

• Sources  of  Information  utilized  by  participants  to  acquire  information  regarding  gluten   were  identified  through  the  question  where  do  you  get  your  information  about  gluten.   Answer  choices  in  this  question  were  categorized  into  two  groups:  medical  versus  social   sources  and  were  examined  using  bivariate  correlation  analysis  (Table  6).    


• Purchasing  behaviors  of  consumers  was  identified  through  the  question  what  prompts   you  to  buy  gluten-­‐free  products.  Answers  to  this  question  were  divided  into  three   categories:  food  allergy,  health  and  nutrition  and  “other”.  The  mean  and  standard   deviation  were  obtained  by  running  frequencies  and  cross-­‐tabulations  (Table  7).      

• Important  features  of  the  food  label  were  compared  between  the  medical  diagnosing   and  self-­‐diagnosis  group  using  a  univariate/two-­‐way  between  groups  analysis  of  

variance  (ANOVA).  The  mean  and  standard  deviation  is  illustrated  in  Table  8,  and  the  full   table  is  displayed  in  Appendix  C.    

• Dietary  Behaviors  of  consumers  were  measured  by  first  looking  at  the  dieting  history  of   the  participants,  which  was  scored  based  on  answers  to  the  questions  do  you  follow  a   special  diet  and  have  you  ever  tried  any  of  the  following  diets.    

! A  score  of  zero  meant  that  the  participants  have  never  followed  a  special  diet.   ! A  score  of  1  meant  that  participants  have  followed  one  diet  in  their  lifetime.   ! A  score  of  2-­‐4  meant  that  participants  have  tried  several  diets  throughout  their  


Dieting  history  means  were  compared  using  an  independent-­‐samples  t-­‐test  and  a  Mann-­‐ Whitney  U  test  for  statistical  significance  reinforcement  (Table  9).    


• Another  set  of  questions  aimed  at  consumers’  strategies  for  using  alternatives  to  gluten-­‐ containing  foods  were  asked  to  measure  dietary  changes  of  participants.  The  answer   options  for  these  questions  included:  any  item  labeled  gluten-­‐free,  only  specific  gluten-­‐ free  brand  names,  other  grains  such  as  corn  and  rice,  meat  products,  dairy  products  or   fruits  and  vegetables.  Answers  to  these  questions  were  compared  to  individuals  who   were  diagnosed  by  a  medical  doctor  and  those  who  self-­‐diagnosed.    Mean  and  standard   deviation  were  obtained  through  frequencies  and  cross-­‐tabulations  (Table  10).    


• Nutritional  consequences  of  a  gluten-­‐free  diet  were  assessed  through  the  question  How   have  the  following  food  groups  changed  in  your  diet  since  starting  a  gluten-­‐free  diet?   (Q27).  An  independent-­‐samples  t-­‐test  was  utilized  to  compare  the  means  of  each  

category  and  a  Mann-­‐Whitney  U  was  used  to  reinforce  any  statistical  significance  finding   (Table  11).    


• Total  adherence  to  a  gluten-­‐free  diet  was  computed  by  combining  the  scores  of  the   following  four  questions:    

! Question  16:  How  often  do  you  buy  gluten-­‐free  products?  

! Question  17:  When  I  do  buy  one  of  the  following  products,  I  buy  the  gluten-­‐free   version:  

! Question  19:  Have  you  had  difficulty  accepting  a  gluten-­‐free  diet?   ! Question  22:  Do  you  ALWAYS  follow  a  gluten-­‐free  diet?  


Question  sixteen  answer  =  “Always”:                            1    +   Question  seventeen  answer  =  “Yes  to  all”:          9    +   Question  nineteen  answer  =  “No,  not  at  all”:  1    +   Question  twenty-­‐two  answer  =  “yes”:                          1  


Total  Score:  12  (100)%      

Total  adherence  score  was  calculated  out  of  twelve  points.  Participants  who  earned  all   twelve  points  received  a  score  of  100%.  The  mean  scores  were  compared  using  an   independent-­‐samples  t-­‐test  and  Mann  Whitney  U  test  was  utilized  to  reinforce   statistical  significance  (Table  12).    


Adherence  to  a  gluten-­‐free  diet  was  furthered  analyzed  through  a  one-­‐way  analysis  of   variance  (ANOVA)  to  identify  statistical  significance  within  the  three  main  medical  

diagnosis  categories  (CD,  GS,  IBS).  A  Post  Hoc  test  was  used  to  identify  the  statistical   significance  between  Celiac  Disease  and  Irritable  Bowel  Syndrome  (Table  13).      

Further,  quality  of  life,  which  is  often  an  indicator  of  adherence,  was  scored  based  on   three  questions.    

! Question  19:  Have  you  had  difficulty  accepting  a  gluten-­‐free  diet?  

! Question  24:  Do  you  feel  embarrassed  about  asking  for  gluten-­‐free  foods  when   you  are  out?  

! Question:  25:  Have  you  given  up  any  group  activities  because  you  are  afraid  you   might  be  tempted  not  to  follow  a  gluten-­‐free  diet?  


Question  nineteen  answer  “Yes”:  1   Question  twenty-­‐four  answer  “yes”:  1   Question  twenty-­‐five  answer  “yes”  :1  


      Total  Score:  3  (100%)    

The  total  score  for  quality  of  life  had  an  inverse  relationship  meaning:   ! Score  0  =  Excellent  quality  of  life  

! Score  1  =  Good  quality  of  life   ! Score  2  =  Average  quality  of  life   ! Score  3  =  Poor  quality  of  life  


Mean  and  standard  deviation  for  this  category  was  obtained  using  frequencies  and   cross-­‐tabulation  (Table  12).    




Two  hundred  and  forty-­‐nine  responses  were  collected.  Thirty-­‐one  respondents  did  not  

complete  the  survey  and  their  responses  were  excluded  from  the  statistical  analysis  resulting  in   an  87.5%  questionnaire  completion  rate.  The  remaining  two  hundred  and  eighteen  responses   were  entered  into  SPSS  for  analysis;  however,  of  all  the  participants  who  did  complete  the   survey  only  respondents  who  identified  as  having  followed  a  gluten-­‐free  diet  (n=73)  were   utilized  for  a  full  analysis.    


In document Knowledge and Behaviors Surrounding a Gluten-Free Diet Between Medically and Self-Diagnosed Individuals (Page 51-57)