Live blogged from Nashville. Any errors or bad jokes are my own.
– We want surveys short and simple. to avoid straightlining, and satisficing. reuce breakoffs, and dropping off the panel.
– but companies are ok with panelists taking multiples surveys in a row
– is multiple short surveys better than one long survey?, assume it lets people handle fatigue better, assumes if they do take another survey that that survey will be better quality. is any of this true?
– who takes multiple surveys, what are their completion rates, how good is the data, how does it affect attrition
– defined surveys as all the surveys taken within 1.25 hours
– 40% of surveys are completed in chains
– younger people make more use of chains
– moderate chaining is the norm. most people average 1.5 to 3 surveys per session. about 10% average more than 3 surveys per chain.
– completion rates increase with each survey in the chain. people who want to drop already dropped out.
– buying rate is unaffected by chaining. for people who take five surveys, buying rate increases with each survey.
– why is this? panelists will take more surveys if they did not exhaust themselves in the previous survey. or maybe those with lots of buying behaviours pace their reporting. or those people are truly different. [read the paper. it’s getting too detailed for me to blog on]
– poor responders are more likely to chain, but not massively more likely
– for younger panelists, heavy chainers have greater longevity. for oldest panelists, it results in burnout.
– people who agree to chain, do it because they are ready to do so. if they exhausted in a previous survey, they don’t continue. a small minority abuse the process
– chaining helps younger panelists stay engaged