Evaluation of role of anxiety
-methodological issues: Johnson and Scott's research lacks mudane realism as PP knew they were part of a study so may have paid more attention to details. Can't generalise findings to real life.
Christianson and Hubinett - high ecological validity as it was a real crime so EWT is more generalisable. However there is the issue of proximity being a confounding variable.
+real life studies increases ecological validity so can be generalised to real life. Real experienced stress and anxiety so more realistic and accurate results from EWT.