Expectations are routinely utilized by the cognitive system to facilitate processing and optimize responses. However, as we age, the access and use of expectations might undergo significant changes. In an eye-tracking study, we investigate how search performance in young and old age is influenced by contextual priming of expectations, and the contextual consistency of the target object. We demonstrate that older participants congruently primed perform significantly worse, especially when the target object is inconsistent with the scene context (e.g., exposed to a 'restaurant' scene before performing the search in another 'restaurant' scene, but looking for an 'iron' rather than 'bread'). Even though congruency and consistency differentially impact on the eye-movement responses of the two age groups, they do not display any interaction. Overall, this study suggests that reliance on mechanisms of contextual expectations becomes stronger when we get older.
Plates and forks are more likely to be found in kitchens than
in bathrooms. Expectations of this sort are constantly
generated by our cognitive system to predictively anticipate the
state of events, and pro-actively generate appropriate
responses
Visual attention is a cognitive modality that relies heavily
on such contextual expectations to maximize performance in
visual tasks
However, as we age, the visual system deteriorates and
performance on visual tasks starts to decrease
Most of the above studies on aging focused on low-level
features, with few other studies approaching this topic from
a naturalistic perspective, where knowledge-based rather
than stimulus-based processing is examined
Priming paradigms are used to probe, and activate, semantic representations in memory. For example, if a prime word is semantically related to a target word (plant-grass), participants' lexical decision responses are shorter than when the prime word does not belong to the same semantic category (Stanovich & West, 1981).
Effects of priming manifest also during visual attention.
The main goal of the current study is to establish how contextual expectations are utilized by the visual system by younger (under 25 years) and senior (over 65 years) participants during a search task. In particular, using a priming paradigm, we probe the congruency of contextual information between two scenes (the prime scene and the search scene), sequentially presented (e.g., bathroom-bathroom, versus, kitchen-bathroom), and the consistency of the target object with the search scene (e.g., a toothpaste versus a fork). In line with research showing that seniors have difficulties with inconsistent cues, we expect them to display a significantly worst search performance when the target object is inconsistent with the primed information. Their difficulties with inhibition suppression would imply an increased difficulty to detect a target object when it inconsistently deviates from the congruently primed contextual expectations. In this eye-tracking study, we compare the behavioural performance (accuracy and response time), and attentional responses (search latency, first fixation and total gaze) of younger and older participants in a visual search task situated in naturalistic scenes. We investigate how contextual priming and the semantic consistency of the search target mediates performance during aging.
Fifty-four participants Portuguese native speakers, divided into two age groups: 32 younger (28 women), age 22.63 (SD = 9.03), University students; and 22 older volunteers (19 women), age 67.9 (SD = 4.75), mostly recruited from senior universities, with no medical history of neuropsychological illness Both groups had normal or corrected to normal vision. The study was approved by the Ethics Committee of the Department of Psychology, University of Lisbon (FPUL).
All participants underwent the Montreal Cognitive Assessment Test - Portuguese Version (MoCA) (Simões et al., 2008), used as a criteria to detect potential mild cognitive deficit, especially for the older group. The MoCA was applied on paper with instructions given orally. The young group scored an average of 27.1 out of 30 on the MoCA (SD = 1.83), while the elderly group scored an average of 24 out of 30 (SD = 2.81). The groups significantly differ in their performance, t(32.8) = 4.46, p < 0.001), however, none of the participants in the elderly group performed below the cut-off score of 19.
Our 2x2x2 experimental design crosses object Consistency (Consistent or Inconsistent), contextual Congruency (Congruent, Incongruent) between the prime scene and search scene, and a between-participants age Group condition (younger or older participant). Refer to Figure 1 for an example of the design material and Figure 2 for the trial run. For example, a participant could be primed with a Restaurant scene, then presented with another Restaurant scene, and previously asked to search for a basket of bread. This example corresponds to the Congruent and Consistent experimental condition.
We utilize 42 experimental scenes, 42 filler scenes and 84 prime scenes, which are all naturalistic, real-world photographs. In particular, the 42 experimental items were drawn from 7 different indoor scenarios (Restaurant, Church, Office, Waiting Room, Bathroom, Living Room, and Kitchen) with 6 items per scenario. The scenes were captured in Lisbon, and surrounding areas, with a congruent or an incongruent object placed by the photographer (refer to section Validity pre-tests of experimental material for details on the validation of the consistency between the target objects and the search scene). The position of the target object was counterbalanced with left-right conditions to avoid the development of search strategies. We decided to take photographs of the scenes to obtain a naturalistic visual saliency, and avoid uncontrolled artifacts (e.g., brightness or contrast) that could be introduced when the target object is digitally inserted into the scene. Material was distributed in 4 lists using a Latin-Square Design.
Primes and fillers were selected from various freely available online databases, such as the SUN database (Xiao, Hays, Ehinger, Oliva, & Torralba, 2010) or Flickr. The primes were selected on the basis of their Congruency with respect to the experimental scenes (refer to next section). Moreover, we made sure that the prime scene never contained the target object. The same databases were used for the filler scenes, and we have chosen scenes belonging to similar contexts. No filler trials did included the target object.
Firstly, we asked 10 students of the FPUL (6 women) between the ages of 20-25 years (mean = 21.5, SD = 1.433) to name 8 objects they considered typical of each of the 7 contexts. 198 unique objects were produced overall, and among those we selected the 8 most nominated objects for each context (e.g., pillow for context bedroom), as the consistent targets. The 8 inconsistent objects, instead, were drawn from another pool of objects never mentioned by any of the participants. A new group of 20 participants (12 women, mean age = 21.2, SD = 1.23), was then asked to judge the consistency between the target objects previously selected (8 consistent, 8 inconsistent) and the embedding context on a 6-point Likert scale (1 – Totally inconsistent; 6 – Totally consistent). We selected the 6 objects that had highest (consistent) and lowest (inconsistent) scores for each context (e.g., Soap – Bathroom: mean = 5.9, SD = 0.31; Flashlight-Bathroom: mean = 1.3, SD = 0.67). Finally, a new group of 12 participants (6 women, mean age = 22.5, SD = 1.87) was asked to respond to another Likert scale questionnaire and rate: (a) the contextual congruency between the prime scene and the search scene, (b) the typicality of the prime scene and search scene given their context, (c) the consistency of the target object within the search scene, and (d) the ease of recognizing the target object given its linguistic denotation.
The purpose of this last questionnaire was to empirically validate the Congruency and Consistency manipulations. Results show that: (a) the difference between congruous prime scene/search scene (mean = 5.4, SD = 1.09) and their incongruent version (mean = 1.5, SD = 0.92), was statistically significant (t(489) = 44, p < 0.001); (b) both prime scene (mean = 5, SD = 0.87) and search scene (mean = 5 , SD = 0.86) were considered to be highly representative examples of their context (t(503) = 0.167, p = 0.87); (c) the target objects consistent with the search scene (mean = 5.1, SD = 1.6) were rated as significantly more likely than their inconsistent counterparts (mean = 1.7, SD = 1.2) t(470) = 28, p < 0.001); and (d) the target objects were rated as highly identifiable by their linguistic denotation (mean = 5.3 SD = 1.7).
Visual stimuli were presented on a LG Flatron L194ws 19inch LCD screen with a 60 Hz refresh rate, 55cm away from participants' eyes. Eye-movements were recorded using an SMI IVIEW X™ HI-SPEED eye-tracker at a sampling rate of 1250 Hz on a 21” screen (1024 x 768 px. image resolution). Viewing was binocular but only the participant’s dominant eye was tracked (determined by a prior parallax test). Chin rest, and forehead support were used to stabilize the head position, and keep viewing distance constant. Participants logged their answers regarding the presence (or not) of the target object in the scene, using a Logitech Cordless Rumblepad 2 controller. They used 3 buttons for: “Yes”, “No” and to advance to the next trial. The experimental protocol was controlled by ExperimentCenter 3.2 (SMI).
Participants were first presented with the target object (a
written cue for 500ms). Then, a fixation cross appeared in
the center of the screen for 1000ms after which the prime
scene is presented for 250ms, followed by a mask2 of the
same scene for 250ms. The mask is utilized to disrupt
feedback between lower and higher visual areas and keep the
perception of the prime scene on a subliminal level
Each block contained 42 experimental trials and 42 filler trials, for a total of 84 trials, which were randomly presented to each participant. The target object was present in the experimental trials, and absent in the filler items guaranteeing a balanced 50-50 distribution of yes/no responses, as well as keeping the participants engaged in the search task. 4 practice trials were administered at the beginning of the experimental session to familiarize participants with the task. A 15-points calibration procedure was performed at the beginning of the trial; and repeated every 10 trials (8-points) to ensure accurate tracking. The calibration was accepted when angle threshold was ≈ 0.8/1.2 for x and y respectively. On average, the younger population had 0.61º ± 0.42º deviation error on the x-axis. and 0.72º ± 0.58º on the y-axis. The older population had 0.82º ± 0.60º deviation error on the xaxis and 1.12º ± 0.79º on the y-axis. The experiment was explained using written instructions and the eye-tracking session lasted for approximately twenty to thirty minutes.
The performance measures analyzed are: (a) response
accuracy, a binomial variable for correct and incorrect
responses (1, 0), and on correct trials only, (b) the response
time to pass to the next trial (in ms.). From the
eye-movement responses to the target object, only for correct trials
2The masks were generated by splitting the prime scene into 12
by 9 quadrants, 100x100 pixel squares, and randomly shuffle them.
we compute: (c) the search latency in ms. (i.e., the time
from scene onset until the target object is looked at for the
first time), (d) the duration of the first fixation in ms., and
(e) the total duration of fixations on target-object in ms.
These eye-movement measures are commonly used to
characterize attentional performance in search tasks
In Table 1, we report means and standard deviation for each dependent measure across all experimental conditions, which are also visualized in Figure 3 and 4. We observe a significant main effect of Congruency, whereby prime scenes congruent with search scenes improve the detection of the target object [F (1, 1153) = 10.2, p < 0.01]. Younger participants are overall more accurate than older participants [F (1,1153) = 44.4, p < 0.001]. Moreover, we find a significant two-ways interaction Consistency:Group, whereby older participants detect more accurately inconsistent than consistent object [F (1,1153) = 4.22 p < 0.05] Moreover, we observe a significant three ways interaction Congruency:Consistency:Group, whereby response accuracy of older participants drastically drops when prime scene and search scene are Congruent, but the target object is Inconsistent with the search scene [F (1,1153) = 14.07, p < 0.001].
Scenes where the target object was consistent with the context of the search scene took longer to be responded [F (1, 1109) = 17.83, p < 0.0001]. Older participants took longer to respond than younger participants [F (1, 1109) = 309.82, p < 0.0001]. When looking at the two-ways interaction, we find that a search scene containing a Consistent target object, which was Congruently primed [F (1, 1109) = 8.98, p < 0.01] yielded faster response times than a scene with an Inconsistent target object; especially in older participants [F (1, 1109) = 6.77, p < 0.01].
Young participants are overall faster than older participants to look at target object for the first time [F (1, 1109) = 95.81, p < 0.0001]. Search latency is faster when the object is Inconsistent with the search scene, compared to when it is Consistent [F (1, 1109) = 24.06, p < 0.001]. However, when prime scene and target scene are Congruent, and the target object is Consistent, we observe significantly shorter search latencies [F (1, 1109) = 4.19, p < 0.05].
On the duration of the first fixation, we only find that: (1) consistent target objects are fixated the first time for longer than inconsistent objects [F (1, 1109) = 4.42, p < 0.05]; and (2) older participants linger on the target object for longer than younger participants [F (1, 1109) = 5.05, p < 0.05].
On the total fixation duration to the target object, we find main effects of Consistency, whereby participants fixated an Inconsistent target object for longer than when it was Consistent [F (1, 1109) = 9.10, p < 0.01]; and of Group, with younger participants displaying shorter total duration of fixation than older participants [F (1, 1109) = 32.96, p < 0.001]. Furthermore, a two-way interaction effect between these two variables was also observed, whereby older participants fixated the target object overall less when it was Consistent with the scene [F (1, 1109) = 19.92 p < 0.001].
Both age groups are clearly influenced by contextual
priming, object consistency, as well as by their interaction.
In particular, correct detections of the target object strongly
depend on the type of priming participants receive.
Incongruent primes improve target detection, but this is
more strongly the case when the object is also inconsistent.
By probing an unrelated contextual template, participants
may become more aware of inconsistency within the search
scene. The most interesting result, however, is that the
detection performance of older participants abruptly drops
when prime and search scenes are congruent, but the target
object is inconsistent with the search scene. This result
confirms previous literature showing that older adults have
more difficulties to inhibit invalid cued information
Expectations are routinely used by the cognitive system to
anticipate information about real-word events, and optimize
appropriate action-responses
In the current study, we used a contextual priming paradigm to probe expectations, and compared visual performance in a search task between younger and older observers. We demonstrated that older observers rely more strongly on consolidated expectations than younger observers. In particular, when the prime and search scene overlapped in context, older participants had a worse performance than the younger, especially if the target object was inconsistent with the primed expectations. This result suggests that older observers might have more difficulties re-computing the semantics of individual objects when primed, and hence compensate for unexpected information. Nevertheless, eye-movement responses did not show a reliable difference in this interaction between congruency and consistency due to aging. Both groups experienced better search performance when the target object was consistent with expectations, but no difference was observed in this interaction due to age.
This study opens intriguing questions on the role of primed contextual expectations in clinical populations affected by neuro-degenerative disease (e.g., Alzheimer); and especially on how would behavioural patterns compare with a normally aging population.
Overall, this work confirms the important role that contextual information plays on the active allocation of visual attention and shows that reliance on it may become stronger as we age.
Fundação Amadeu Dias for student award to MB, and Fundação para a Ciência e Tecnologia research fellowship