Message words. An independent groups t-test was used to assess if the words in the physical messages differed from the words in the social messages in valence (i.e., negative vs. positive) ratings. A significant difference was found between the physical (M = 3.97, SD
= 1.23) and the social message word valence ratings (M = 5.01, SD = 0.90), t(24) = -2.45, p = .022, 95% CI [-1.90, -0.16], 2 = .11. These findings suggest that participants rated the words in the physical messages as more negative than the words in the social messages (which were rated, on average, as more positive). As such, messages were revised accordingly (refer to the discussion, section 6.7 for details).
Filler words. Paired t-tests were used to examine the difference in valence ratings between the physical message and corresponding filler words and between the social message words and corresponding filler words. There were no significant valence differences between the physical message words (M = 3.97, SD = 1.23) and physical filler words (M = 4.19, SD = 0.85), t(12) = -0.83, p = .421, 95% CI [-0.76, 0.34], 2 = .16, nor between the social message words (M = 5.01, SD = 0.90) and social filler words (M = 4.76, SD = 0.94), t(12) = 1.17, p = .265, 95% CI [-0.23, 0.73], 2 = .16. Thus, the physical and social message words and their corresponding filler words were matched on valence, as intended. Tables 6.2 and 6.3 present the word ratings and other characteristics for the physical and social message words and corresponding filler words, respectively.
89 Table 6.2
Structural Characteristics and Participant Perceptions of the Physical Words and Corresponding Filler Words (N = 24)
Physical message words Corresponding filler words
Word Length Frequency Arousal Valence Word Length Frequency Arousal Valence
each 4 253.25 1.67 4.08 easy 4 267.71 2.46 5.13
year 4 277.92 1.63 3.83 hand 4 279.65 2.08 4.17
australia 9 2.63 3.13 5.08 singapore 9 2.24 2.21 4.33
people 6 1102.98 3.13 4.42 father 6 554.49 3.46 4.63
obey 4 8.94 2.71 3.58 rank 4 8.49 2.25 3.75
one 3 3072.24 2.21 4.17 any 3 1099.37 1.75 4.13
posted 6 7.20 1.63 4.08 lounge 6 7.86 2.42 4.79
physical 8 27.18 3.63 4.13 confused 8 32.41 2.88 2.63
sustain 7 2.67 3.04 4.04 inspect 7 2.55 2.67 4.09
loved* 5 110.33 4.75 6.38 smart* 5 96.25 3.38 5.88
crash* 5 28.65 4.79 1.42 alert* 5 20.61 4.00 3.79
protect 7 70.25 - - forgive 7 70.25 - -
event* 5 26.37 3.04 4.46 actor* 5 26.33 2.67 4.13
chance 6 241.24 - - change 6 240.34 - -
severity* 8 0.59 4.33 2.00 swapping* 8 0.76 2.22 3.83
number 6 240.94 - - dinner 6 202.67 - -
Total Means 5.81 342.09 3.05 3.97 Total Means 5.81 182.00 2.65 4.25
Note. Arousal scale (1 = low arousal, 7 = high arousal); Valence scale (1 = negative, 7 = positive). The words, ‘protect’, ‘chance’, and ‘number’
in the message word list and the words, ‘forgive’, ‘change’, and ‘dinner’ in the filler word list were not evaluated in terms of word valence and word arousal in Study 1a. ‘*’ not included in the final word list. Section 6.7 discusses this word list in further detail.
90 Table 6.3
Structural Characteristics and Participant Perceptions of the Social Words and Corresponding Filler Words (N = 27)
Social message words Corresponding filler words
Word Length Frequency Arousal Valence Word Length Frequency Arousal Valence
choose 6 48.06 2.59 3.96 decide 6 50.14 2.85 4.03
friend 6 419.29 5.11 6.56 family 6 354.25 5.27 6.44
car 3 483.06 3.85 4.74 dad 3 507.25 4.59 5.78
showing 7 31.12 2.96 4.11 weather 7 34.24 2.96 4.11
really 6 1500.16 2.74 4.30 little 6 1446.39 2.00 3.37
care 4 485.25 4.56 5.70 hope 4 320.63 4.37 5.81
safety 6 32.33 4.44 5.78 spring 6 31.31 3.44 5.00
feel 4 627.24 4.70 4.85 must 4 699.24 2.82 3.74
comfortable 11 47.22 4.22 5.66 responsible 11 45.06 4.81 5.55
confident* 9 10.62 4.44 5.96 automatic* 9 6.98 3.22 4.52
put 3 828.45 2.19 3.63 ask 3 483.14 2.63 3.96
being 5 485.90 3.41 4.48 woman 5 434.63 3.70 5.29
best 4 404.37 3.93 5.37 real 4 442.80 3.14 4.33
Total Means 5.69 415.62 3.78 5.01 Total Means 5.69 373.54 3.52 4.76
Note. Arousal scale (1 = low arousal, 7 = high arousal); Valence scale (1 = negative, 7 = positive). ‘*’ not included in the final word list.
91 6.7 Discussion
The first aim of Study 1a was to assess the validity of the manipulation of message frame. The results indicated that participants perceived the gain-framed messages to be more positive and the physical loss-framed messages to be more negative. As such, these messages were perceived by the participants as intended. However, the social loss-framed message was rated to be neutral by the participants. Further, Study 1a aimed to assess the arousal and valence ratings of the individual message and filler words. While words in the physical and social messages were matched on arousal, the findings also indicated that the physical words were perceived to be slightly more negative compared to the social words (by 1 point on a 7-point scale).
To control for this difference in perceived valence between the physical and social messages word lists in the subsequent two studies, changes were made to the physical message and word stimuli. Specifically, two words were changed in the physical messages (“risk” changed to “chance” and “severity” changed to “number”) and an additional two words were changed in the physical message word stimuli list (“event” was replaced with
“chance” and “crash was replaced with “protect”).22 As a result of these word changes in the physical messages, the mean word valence for these message word lists should be higher (i.e., more positive) and therefore, be more likely to match the average word valence rating of the social messages. To validate these word changes made to the physical messages and to the physical word stimuli list, an additional word valence manipulation check was included in Study 2 (the results from Study 2 indicated that the words included in the social and physical messages were matched on valence. See chapter 7, section 7.5.4). Finally, the filler words
22 Despite the word changes in the physical messages and word list, similar word frequencies were still obtained for the physical and social message conditions.
92 that best matched the message words in terms of perceived arousal and valence ratings were
retained for inclusion in the LDT in Study 2.
93 Study 1b: Piloting of the motor vehicle message
6.8 Method 6.8.1 Participants
A total of 21 participants were recruited via email and online advertisements. Of those participants, 18 were undergraduate students and three were recruited from the local
community. Selection criteria included being between 17 and 25 years of age and holding a current provisional or open Australian driver’s licence (8 held an open licence, 13 provisional restricted licence). The participants (17 female) had a mean age of 20.06 years (SD = 2.40).
Seventeen participants listed English as their first language. Of the 21 participants, 14 (66.7%) reported regularly driving over the posted speed limit. The undergraduate
participants were offered partial course credit of 0.5% for their time. All other participants were offered a chance to receive one AUD$50 shopping gift card.
6.8.2 Design
Paired t-tests assessed participants’ valence and arousal ratings of the words included in the motor vehicle message. Independent groups t-tests were used to compare the word arousal and valence ratings between the motor vehicle message and the social messages (from Study 1a), which were used in Study 2 to activate the BIS.