No announcement yet.
  • Filter
  • Time
  • Show
Clear All
new posts

  • Contingency tables with missing values but chi2 with valid values

    Hi everyone,
    I'am comparing a dichotomic variable between 2 groups. Group n°1 is 199 observations including 2 missing values and Group n°2 is 110 observations including 1 missing value. I want to tabulate percentage based on the hole sample per comparison group so I will have a fix n for each group but present chi square statistics based only on valid values. Does it make sense? I wonder if this correct...If not what alternative do I have?
    Thank you

  • #2
    Please clarify with an example.
    If this is the situation
    Do you smoke cigarettes? Yes / No

    Group n1 is Urban - 199 obs with2 missing values
    Group n2 is Rural - 110 obs with 110 obs with 1 missing values

    group n1 group n2
    yes no
    no yes
    no no
    yes yes

    Is this your questions problem?