Räkna med det interna bortfallet [Count with the internal drop-out]

• Missing data is a pragmatic factthat must be investigated and nota disaster to be mitigated. It is aproperty of the population towhich we seek to generalize andcan cause problems not onlythrough its impact on the samplesize available for analysis butalso through its potential hiddenbiases. Making imputations withoutfirst analysing the randomnessof the missing responses caneven be worse than doing nothing,so care is needed whileimputing missing values. Thispaper reflects on how to prevent,analyse and handle missing dataand how effects of imputation canbe checked.

Keyword

Statistics

missing data

drop-out

methodology

caring sciences

Publication and Content Type

vet (subject category)

art (subject category)