Long-term outcome of cognitive therapy's contribution to self-exposure in vivo to the treatment of generalized social phobia.

Bennett et al. (1998) asked a simple question: does adding extra parts to exposure therapy help adults with social phobia?

They split adults with generalized social phobia into three groups. All groups faced real-life scary situations. One group got only exposure. A second group added cognitive therapy. A third group added a self-help book.

The team checked everyone again one year later to see who still felt better.

One year on, about two-thirds of all patients felt less fear.

Surprise: the extra parts did nothing. Cognitive therapy and the book did not beat plain exposure.

More parts did not make better long-term results.

Wuang et al. (2012) and Bhaumik et al. (2009) look like they disagree. They found virtual-reality CBT beat waiting lists. The key gap: they had no plain exposure group. Their VR package matched the exposure-only arm in K et al., so both lines agree that facing fear works.

Arwert et al. (2020) updated the story. They used newer trans-diagnostic CBT and saw gains in quality of life. Their program added life-skills modules, something K et al. did not test. The newer work shows you can get extra benefit, but only if you change the whole protocol, not just bolt on cognitive talk.

Halford et al. (2002) backs the same theme. They showed that even adults packed with negative core beliefs still gain from standard CBT. Again, no need for extra cognitive add-ons.

If you run exposure for social anxiety, keep it simple. Bring clients into real situations, reinforce approach, and track fear drops. Skip extra worksheets or thought logs unless the case truly needs them. You save hours and still get the same one-year payoff.