Signal detection comparisons of phonemic and phonetic priming: The flexible-bias problem

Abstract

The phonemic priming effect may reflect the hidden dynamics of spoken word perception and has thus been a key topic of recent research. This investigation compared phonemic and phonetic priming (cf. Goldinger, Luce, Pisoni, & Marcario, 1992), using signal detection methods. Although these methods were intended to provide separate indices of sensitivity and bias changes, the results were more complex. Instead, phonemic priming engendered a flexible, trial-specific strategy that affected hits and false alarms (and thereby altered sensitivity) but also created behavioral changes indicative of a bias. Together with previous research, the results suggest that phonemic priming data must be interpreted with caution, and they underscore the limitations of signal detection analyses in priming research (Norris, 1995). However, if a researcher can anticipate the likely form a bias will assume, signal detection methods can reveal priming effects.Relative to the voluminous literature on semantic priming effects in printed and spoken word perception, formbased priming has received little attention. Form-based priming refers to experiments in which primes and targets are related along physical, rather than conceptual, dimensions.

Revisiting bias effects in word-initial phonological priming.

“…The best course of action that one could take would be to incorporate into the experimental design manipulations that counteract such strategies. Goldinger (1998Goldinger ( , 1999) suggests a few possibilities.…”

“…In two studies using multiple tasks but not shadowing, Goldinger et al (1992; see also Goldinger, 1998) found converging evidence that the RT advantage in the 1-and 2-overlap conditions is due to response bias. One of the most convincing demonstrations of this came from a series of lexical decision experiments in which the interstimulus interval (ISI) between prime and target and the percentage of trials on which the prime and target were phonologically related (PRP) were varied, two manipulations effective in past work at reducing the influence of response biases (Neely, 1977(Neely, , 1991Posner & Snyder, 1975).…”

“…The purpose of the present study was to help resolve this dispute. Our approach combined a few of the methods of Goldinger (1998). Whereas Goldinger focused only on the RT costs (i.e., slowdown) in the 0-overlap trials over the course of the experiment as a measure of response biases, we also measured corresponding changes in the 3-overlap condition.…”

Bias effects in facilitatory phonological priming

“…Subjects appear to learn that target words are likely to begin with the same segments as their primes; hence, they prepare the production of those segments and, thus, repeat the targets more rapidly (Goldinger, 1999;Hamburger & Slowiaczek, 1999). The effect is not observed in the lexical decision task (Radeau, Morais, & Dewier, 1989;Slowiaczek & Hamburger, 1992;Slowiaczek & Pisoni, 1986), except when materials are presented in noise and a relativelyhigh proportion of related trials is included (Goldinger, 1998b;Goldinger, Luce, Pisoni, & Marcario, 1992). Again, Goldinger et al (1992) explained the facilitation they observed in terms of bias: Expectations of shared initial segments between primes and targets could benefit performance on targets in noise even when no naming response is required.…”

“…It is much less clear how such a bias would influence shadowing performance, where the naming response varies from trial to trial. If the bias acted at the response stage, influencing the speech production processes required for a naming response, it would have to be flexible, in the same sense as the strategy proposed by Goldinger (1998b) for single-phoneme onset overlap facilitation. That is, the bias would need to be reapplied on each trial, so that the encoding of the phonologicalmaterial for each particular target word could be benefited in some way by the presence of rhyme overlap.…”