Surveys are notoriously fickle when it comes to wording. A slight wording change can change a respondent’s answer. As such, data analysts need to stick very, very close to the survey question and response category wording when reporting on findings to avoid inaccuracies in reporting.
.
Real World Example
A remarkable news headline: “95 per cent of young people are struggling with their mental health.”
Does this ring true to you? Almost every young person is struggling with their mental health? It didn’t ring true to me either, so I did a little digging into the source survey. ARTICLE LINK1 ARTICLE LINK2 SURVEY FINDINGS
The survey found that 95% of people age 18-24 had experienced anxiety at some point in the past year. It is this finding that has been paraphrased into “95% are struggling with their mental health.” The survey didn’t ask how many people today are experiencing anxiety, nor whether they
struggled” with anxiety. At this stage in life, young people are transitioning from school to work or higher education, and often moving out on their own for the first time. A bit of anxiety occasionally is normal. The survey did not find that 95% of young people are currently experiencing anxiety nor that they had been “struggling” with anxiety.
The survey did find, however, that 21% of young adults had experienced anxiety “all the time” in the past 12 months. Surely this finding is worrisome enough to merit headlines without exaggerating (falsifying?) their findings.