Six components, two scoring rules

Griffiths’s components model counts a component as present when it is rated 3 or higher on a 0–5 scale, then asks whether the profile is an addiction. A strict monothetic rule demands all six present; a looser polythetic rule demands only a chosen minimum. Set each component and the polythetic minimum, and watch the same profile change classification on the rule alone.

Salience4 (present)
Mood modification4 (present)
Tolerance3 (present)
Withdrawal2 (absent)
Conflict3 (present)
Relapse1 (absent)
Polythetic minimum (components required)4 of 6
SalienceMoodToleranceWithdrawalConflictRelapse
4 of 6 components present. Monothetic (all six): not addiction. Polythetic (at least 4): addiction. The same answers yield opposite verdicts: loosening the rule from all-six to 4-of-six is what reclassifies the profile — a decision made by the researcher, not the respondent. This is the overpathologizing critique in miniature.

The I-PACE loop: escalation or check

In the I-PACE model each use episode that brings relief or gratification strengthens the cue-reactivity that led to it, while inhibitory control works against that growth. Raise the predisposing vulnerability or lower inhibitory control and the loop escalates toward a compulsive fixed point; strengthen control and the same loop is held in check. The curve iterates rn+1 = rn + gain·rn(1 − rn) − control·rn over successive episodes.

Predisposing vulnerability60
Inhibitory control30
0.000.250.500.751.00cue-reactivityuse episodes →
After 24 episodes, cue-reactivity settles at 0.65 — an escalating, compulsive trajectory. Reinforcement outruns control: each episode strengthens the cue response more than inhibition restrains it, and use drifts from deliberate toward compulsive — the gratification-to-compensation shift the model describes.

The cutoff makes the prevalence

Here is one fixed population of 1,000 people and their screening scores — nothing about them changes. All that moves is the cutoff above which a score is called “addiction.” Slide it and watch the headline prevalence swing. A strict cutoff recovers a few percent; a lenient one multiplies it several-fold, from the very same unchanged scores.

Addiction cutoff (score at or above)18
06121824screening score →people
At a cutoff of 18, 88 of 1,000 people score at or above it — an apparent prevalence of 8.8%. The population never changed; only the bar did. This is why a reported prevalence of technology addiction is uninterpretable without the cutoff and scoring rule that produced it — the same phenomenon can be a single-digit rarity or a double-digit “epidemic” by choice of threshold alone.