Table 2.

List of questionnaire answer pattern indices.

Index				Description
Item nonresponse
		Proportion of skipped items [18]		This index measures the proportion of skipped items in a multi-item survey, which has been suggested to be associated with cognitive functioning [18].
Response styles
		Acquiescence (“yea-saying”) Disaquiescence (“nay-saying”) Extreme response tendency Midpoint responses		These indices are based on detecting abnormal response styles to identify respondents with cognitive impairment. Emerging evidence has suggested that certain response styles, such as acquiescence and an extreme response tendency, may be predictive of cognitive impairment or associated with important risk factors for cognitive impairment like low education in older adults [19,20].
Expressions of uncertainty
	Proportion of “don’t know” answers Response uncertainty parameter [21]		These indices capture the extent to which participants express uncertainty about their answers in the surveys.
Classical Test Theory indices of improbable responses
	Multivariate outlier rate [22] Person-total correlation [23] Psychometric synonyms [24] Psychometric antonyms [24]		These indices are based on detecting improbable or implausible responses to identify respondents with cognitive impairment. Cognitive impairment has been suggested to result in improbable answers [24].
Item Response Theory person-fit statistics
	Polytomous lz index [25] Normed Guttman errors [26] U3 index [27]		These indices are also designed to detect improbable or implausible response patterns, based on modern test theory approaches.
Random errors or noise in responses
	Individual test reliability Inter-item standard deviation [28]		These indices capture unreliable answer patterns to capture random variability in attention or fluctuating cognitive performance.
Response inertia
	Residual autocorrelation across consecutive responses [29] Long-string analysis [24]		These indices measure carry-over effects in people’s response patterns, which may express an inability to adapt responses to changing question contents.
Response pattern drift within a survey
	Improvement or worsening in missing, stylistic, outlying, misfitting, or random response patterns within a survey Change-point analysis of weighted residuals [30]		These indices measure trends in response patterns over the course of a survey. Worsening response behaviors as a participant completes multiple questions in succession may indicate mental fatigue or lower sustained attention ability.