Wise question, Gray.

Unless you ask it and answer it with reasonable evidence, tool box talks are at best window-dressing and may be wasteful and harmful.

Toolbox talks, like any other safety intervention, can CERTAINLY 'work' in the sense of achieving agreed objectives PROVIDED
a. you make the objectives clear
b. these objectives are understood and supported by the other people involved
c. you design reliable valid measures of the changes you wish to bring about through the talks
d. you use these measures honestly and consistently.

Because conditions c and d are often challenging, they are commonly omitted with the result that the effectiveness of tool box talks become at best a matter of hope. Where you do use reliable valid measures honestly and consistently, you can not only improve the efficacy of safety management at least threefold but you gather quantified evidence of economic results.

It's best to start simply, with one or two work groups and to use what's known as a 'control' group, i.e. a group similar in makeup that doesn't have the talks. In that way, you have what's called technically a 'quasi-experiment'.

If you need any informal help with step c, you're welcome to have a brief conversation during office hours; I'm at 020 8654 0808.