Stockholm for They found significant effects of ordering whereby those rating personal risk first, and then comparative others’ risk, exhibited lower levels of UO than those receiving the questions in the opposite order. To illustrate ordering effects further, Budd carried out an experiment where the order of theory of reasoned action items was muddled across different versions of a questionnaire. He found that muddling significantly altered the intercorrelations between perceptions of threat, attitude, normative beliefs and intention to either smoke, brush teeth three times per day, or exercise for twenty minutes. Sheeran and Orbell tried to replicate this using protection motivation theory components see later in the chapter in relation to different behaviour ? that of condom use and dental flossing. While fewer effects of ordering were found, correlation strengths between some key cognitive variables did change. These authors also report that scores on a social desirability scale, along with the perceived salience relevance and importance to the individual of the behaviour being addressed, had small but reliable effects on the associations between the health beliefs assessed in the PMT. These types of study highlight the need for researchers to take possible demand effects and questionnaire design factors into consideration when interpreting their findings. Stockholm 2016.