Mixing reversed items can flag careless answers, and it raises the reading load

What it is

Inserting opposite-direction items for the same construct—“this flow is clear” next to “this flow is a mess”—catches people who tick the same side all the way down. That mix of reverse-coded items can detect acquiescence and satisficing. The cost is that readers must switch polarity between items, so working memory and reading load rise, and careful respondents also reverse by mistake. Mixing is a quality-control device, not a free reliability upgrade.

Why it happens

Satisficing strategies include agreeing with everything, sitting on the midpoint, or sweeping in one direction. A reversed item turns those strategies into internal inconsistency, so a threshold can drop the record. Polarity switches, though, require re-parsing negation and antonyms on every item; on a phone, in a second language, when tired or timed, failed parses create a method factor: reversed items cluster as if they were a second dimension when they are a reading style. Reverse sentences written “for balance” are often not mirrors of the forward ones (“not clear” is not “chaotic”), so the construct quietly becomes two nearby things and a drop in correlation is misread as complex attitude.

Studying it

Confirm in cognitive interviews that the reverse sentence and the forward sentence are heard as ends of the same object. In factor analysis, watch whether reversed items form their own factor; if they do, compare a unidimensional model after dropping satisficers with a model that treats reversal as a method factor. Report how many records were dropped for inconsistency and how those people performed on the task, to see whether the rule is dropping struggling users rather than careless ones. A controlled comparison of an all-forward form with a mixed form on burden, time, and criterion correlation beats assuming that mixing is more scientific.

Where it stops holding

With children, low literacy, screen readers, and translated instruments, misreads of reversed items often outrun the quality-control gain; attention checks, minimum times, or repeated items are safer than polarity flips. When the construct itself is a negation (“distrust”), reversing again produces a double negative, which is a wording failure rather than a mixing tactic. On a very short scale, one reversed item can barely separate satisficing from genuine disagreement and is still enough to scramble the total.

Applying it

• When quality control is needed, keep reversed items few, syntactically simple, and strict mirrors of the forward items, with a drop rule written in advance.
• Keep the main scale in one polarity; do not flip half the items to look rigorous.
• Describe dropped records on their own and check for over-dropping on particular devices and languages.
• Check: if the mixed form takes clearly longer than the all-forward form and criterion correlation does not improve, withdraw the mix and use another quality-control method.

Related

• Same group: Q3.15.1 Likert scoring assumes one continuous attitude · Q3.15.2 Option order produces primacy and recency · Q3.15.4 Cross-language scales need back-translation, not a word-for-word rendering
• Adjacent: Q3.01 Questionnaire design · Q1.07 Researcher bias
• Search terms: reverse-coded item · satisficing · acquiescence