Implicit and explicit preferences for brand name sounds

The target words are significantly less familiar and meaningful than the benchmark (M_{target word familiarity} = 1.60, SD = .76 vs. M_{benchmark familiarity} = 3.38, SD = 1.40; t(77) = 12.05, p < .001; M_{target word meaningfulness} = 1.66, SD = .83 vs. M_{benchmark meaningfulness} = 3.32, SD = 1.47; t(77) = 10.69, p < .001). The averages of “S” and “TH” word familiarity are similar (M _{“S” word familiarity} = 1.66, SD = .88 vs. M_{“TH” word familiarity} = 1.50, SD = .74; t(77) = 2.00, p = .049) and “S” and “TH” words are equally unmeaningful (M _{“S” word meaningful} = 1.69, SD = 0.88 vs. M_{“TH” word meaningful} = 1.61, SD = 0.91; t(77) = 1.03, p = .31).

Per Greenwald et al. (2003), trials were dropped if they exceed 10 s, which happened rarely (< 0.01% of trials). Participants were dropped if > 10% of trials in combined blocks were < 300 ms; drop rates using this procedure were acceptable (10%). There was no error penalty because participants were forced to correct errors before moving forward to the next trial. Finally, the IAT was scored using Greenwald’s D-score algorithm, which calculates the difference between each participant’s mean latencies on the two critical blocks, divided by the participant’s inclusive standard deviation, similar to Cohen’s d. Reliability, as measured by a split-half reliability procedure (De Houwer and De Bruycker 2007) with a Spearman-Brown correction (Revelle and Zinbarg 2009) was good (0.80).

Brand name stimuli are significantly less familiar and meaningful than the benchmark (M_{brand name familiarity} = 1.88, SD = 1.04 vs. M_{benchmark familiarity} = 3.09, SD = 1.67; t(139) = 10.95, p < .001; M_{brand name meaningfulness} = 1.74, SD = 1.04 vs. M_{benchmark meaningfulness} = 2.86, SD = 1.52; t(139) = 10.19, p < .001). Importantly, the words in each target pair (e.g., Moddip/Noddip) are equally unfamiliar (all p’s > .30) and unmeaningful (all p’s > .25). Finally, “M” and “N” brands are on average equally unfamiliar (M _{“M” brand familiarity} = 1.83, SD = 1.21 vs. M_{“N” brand familiarity} = 1.86, SD = 1.25; t(122) = − 0.31, p = .75) and unmeaningful (M _{“M” brand meaningful} = 1.72, SD = 1.15 vs. M_{“N” brand meaningful} = 1.65, SD = 1.00; t(122) = 0.86, p = .39).

After observing large effects in studies 1 and 2, the remaining studies used samples of approximately 50, which provide adequate power to detect medium-large effects. Given that our paradigm tests preferences under ideal laboratory conditions, and the effects may be weaker under market conditions, we consider medium-large to be the smallest effect size of interest.

Target stimuli words are significantly less familiar and meaningful than the benchmark (M_{target word familiarity} = 1.89, SD = .97 vs. M_{benchmark familiarity} = 3.47, SD = 1.61; t(80) = 10.41, p < .001; M_{target word meaningfulness} = 1.95, SD = 1.07 vs. M_{benchmark meaningfulness} = 3.50, SD = 1.44; t(80) = 11.22, p < .001). Importantly, the words in each target pair (e.g., Loddip / Woddip) are equally unfamiliar (all p’s > .30) and unmeaningful (all p’s > .15). Also, “L” and “W” words are on average equally unfamiliar (M _{“L” word familiarity} = 1.90, SD = 1.05 vs. M_{“W” word familiarity} = 1.87, SD = 1.18; t(80) = .29, p = .77) and unmeaningful (M _{“L” word meaningful} = 1.99, SD = 1.22 vs. M_{“W” word meaningful} = 1.89, SD = 1.08; t(80) = 1.21, p = .23).

Target stimuli words are significantly less familiar and meaningful than the benchmark (M_{target word familiarity} = 1.85, SD = 1.08 vs. M_{benchmark familiarity} = 4.37, SD = 1.46; t(147) = 19.05, p < .001; M_{target word meaningfulness} = 1.82, SD = 1.06 vs. M_{benchmark meaningfulness} = 4.06, SD = 1.50; t(147) = 16.89, p < .001). Importantly, the words in each target pair (e.g., Seedris/Sudris) are equally unfamiliar (all p’s > .25) and unmeaningful (all p’s > .10). Also, “E” and “U” words are on average equally unfamiliar (M _{“E” word familiarity} = 1.83, SD = 1.27 vs. M_{“U” word familiarity} = 1.81, SD = 1.17; t(135) = .16, p = .87) and unmeaningful (M _{“E” word meaningful} = 1.83, SD = 1.27 vs. M_{“U” word meaningful} = 1.72, SD = 1.05; t(135) = 1.16, p = .25).