Understanding convictions, your own included

Sorting people into named profiles makes a result legible. It also creates an illusion: that of clear frontiers where there are only degrees.

Around age two, a child starts treating a stick as a telephone. That conduct is a solid marker; what it produces is far less established.

French law requires employers to monitor the workload of staff on annual-day contracts. No reference instrument says how.

Neighbour relationships have receded in personal networks since the 1960s. That retreat mainly concerns one precise kind of use.

An innocuous remark triggers in you a disproportionate reaction, always the same, in the same circumstances. That is not an anomaly, it is ordinary learning.

A study covering more than ten thousand couples compared the predictive value of everything research thought it knew. Its conclusions are disappointing, and instructive.

An association between a parenting practice and a child outcome may be produced by neither of them. Recent designs make it possible to test that.

Since 1990, one survey has asked comparable questions in more than thirty countries. It is the field's most solid source, and it has its limits.

Systematic reviews exist and their conclusions are cautious. The main difficulty is not the effectiveness of the schemes but the quality of their evaluation.

A well-built self-description informs. It predicts neither a career, nor a relationship, nor a future reaction, and claiming otherwise is a methodological fault.

We judge our relationship against those we believe we are observing. The problem is that we are not comparing what we think we are comparing.

Not voting is not a single position. It is one and the same act covering very different situations, and conflating them leads to false conclusions.

A concern report and a referral to the prosecutor do not go to the same place and do not have the same effects. The confusion produces unusable figures.

The longer a spell without work lasts, the lower the probability of leaving it. Separating selection from the effect of duration itself required particular designs.

Both events reduce the number of active relationships, but not in the same way nor for the same people.