Abstract
Imagine not being able to ignore the sense of your shirt on your body or the odor of your fabric softener. Usually, the sensory system alerts to new stimuli and then gradually ceases to respond in a process called habituation. Habituation is the capacity to decrease response toward stimuli of different modalities, enabling people to filter out irrelevant stimuli and focus selectively on important ones (Peeke & Petrinovich, 1984).
Some people have slow or impaired habituation, being constantly aware of a particular stimulus; consequently, they may suffer from deficits in daily function (Brown et al., 2001). Deficits in the habituation process affect cognitive functions and correlate with eating disorders, addiction, and chronic pain (e.g., Çevik, 2014; Coppola et al., 2009; De Luca, 2014). This article describes the development and psychometric properties of the Sensory Habituation Questionnaire (S–Hab–Q), designed to characterize habituation to daily sensations, an important dimension of sensory overresponsivity (SOR).
According to Dunn (1997), the balance between habituation and sensitization can explain the ability to respond effectively to the sensory environment. SOR may reflect two dimensions: (1) elevated sensitivity to stimuli at the detection level and (2) decreased habituation to stimuli (Green et al., 2013, 2015; Schipul et al., 2012). The physiological mechanism of habituation relates to an increased inhibitory process in specific brain cells that eventually prevents other brain cells from reacting to the stimulus. This process, called gating, results in a reduction of brain reactivity to repeated stimuli and has been demonstrated in mice and in humans (Kato et al., 2015).
Specific clinical populations are more prone to sensory habituation deficits. For example, evidence from people with autistic spectrum disorder suggests a prolonged processing of certain stimuli (Hudac et al., 2018). Research on fragile X syndrome has shown fixation to auditory stimuli with synaptic hyperexcitability causing decreased habituation (Ethridge et al., 2016). People with Tourette syndrome present with reduced GABAergic transmission, delaying inhibitory processes in response to tactile stimuli (Puts et al., 2015). Decreased sensory habituation has also been identified as a potential mechanism that contributes to the interplay between SOR and psychopathology (Belluscio et al., 2011; Güçlü et al., 2015).
People with psychopathology often report SOR (e.g., Conelea et al., 2014; Lipskaya-Velikovsky et al., 2015), which exacerbates their clinical symptoms (Cohen & Leckman, 1992; James et al., 2011). The task of occupational therapists is to detect SOR and accurately differentiate it from symptoms of psychopathology. However, previous research has demonstrated that measures assessing each construct overlap, which may hamper their differentiation (Ben-Sasson et al., 2007; Gouze et al., 2009).
Sensory questionnaires mostly inquire about the frequency of behavioral and emotional responses toward daily sensory stimuli (Brown & Dunn, 2002), using terms such as “I fear” and “I get frustrated when . . . .” People completing the questionnaires might confuse these responses with other symptoms and experiences not necessarily related to SOR. Thus, when developing the S–Hab–Q, we ensured that the items were free from emotional descriptors, similar to the Sensory Perception Quotient (SPQ) developed to measure sensitization (Tavassoli et al., 2014). The SPQ contains items such as “I would be able to distinguish different people by their smell,” whereas the S–Hab–Q has items such as “I hear the sound of a computer’s CPU while I’m working on it” and “When the light in the room is bright, I can’t read.” We want to provide clinicians with a tool to capture the perceptual nature of sensory habituation while minimizing evaluation of negative emotionality stemming from other sources.
Our main aim in this preliminary study was psychometric examination of the S–Hab–Q in a neurotypical sample. Therefore, we applied the Classical Test Theory (Novick, 1966) approach and the multitrait–multimethod matrix (Campbell & Fiske, 1959) to establish assessment convergent and divergent validity. We developed the following five hypotheses:
The S–Hab–Q will demonstrate satisfactory internal reliability (r > .70; Nunnally & Bernstein, 1967).
The S–Hab–Q and SPQ will show moderate correlations as a result of measuring different aspects of SOR.
The S–Hab–Q and the Adolescent/Adult Sensory Profile (AASP; Brown & Dunn, 2002)–SOR will show moderate correlations; however, these correlations will be smaller than Hypothesis (H) 2 because of the different elements being measured by each tool.
The S–Hab–Q and AASP’s other quadrants will show some correlations as a result of habituation’s contribution to each quadrant.
The S–Hab–Q and the Brief Symptom Inventory (BSI; Derogatis & Melisaratos, 1983) will not correlate because each measure has distinct constructs.
Method
Phase 1: Questionnaire Construction and Content Validity
Thirty-two potential items were selected for the questionnaire on the basis of a literature review (Dunn, 1997; Miller et al., 2001; Peeke & Petrinovich, 1984; Thompson & Spencer, 1966) and existing sensory questionnaires for children and adults (e.g., Brown & Dunn, 2002; Robertson & Simmons, 2013; Schoen et al., 2008). A time-based scale was chosen to reflect the temporal dimension—a basic element of habituation—and a 4-point Likert scale was defined to infer the habituation process: 0 = not long at all, 1 = not a very long time, 2 = an extremely long time, and 3 = I can’t get used to it.
The 32 items were sent to nine international sensory modulation experts, all of whom had developed tools in the sensory field. All but one expert had a PhD and held a senior academic position. For each item, we asked the experts to rate the degree to which they agreed that the item represented sensory habituation. In addition, we asked them to rate the degree to which the item was well phrased and to add their definition of sensory habituation. Last, we asked to what degree the experts thought the S–Hab–Q item pool represented a proper sample of sensory habituation and to suggest important indicators of sensory habituation we had not included. Our inclusion criterion for each item was 77% agreement between experts (Hardesty & Bearden, 2004). Only 7 items did not meet the criterion (Table 1). According to the experts’ suggestions, some items were rephrased to better describe sensory habituation. After this process, we retained 25 items.
S–Hab–Q Initial Items and Experts’ Evaluations
Note. CPU = central processing unit; S–Hab–Q = Sensory Habituation Questionnaire.
Phase 2: Psychometric Property Assessment
Once the final version of the S–Hab–Q was established, we examined its internal reliability and convergent and divergent validity.
Participants
The nonclinical convenience sample consisted of 160 adults (56.87% women, mean [M] age = 31.85 yr, standard deviation [SD] = 10.72). The sample consisted of 88.1% native Israelis and 11.9% immigrants to Israel. Most (60.6%) participants had some higher education (8.1% nonacademic higher education, 35.6% bachelor’s degree, 16.9% master’s degree or doctorate). The rest (39.4%) had only completed high school. Of those reporting marital status, most (54.77%) were single, 42.67% were married, and 2.56% were divorced or widowed. Inclusion criteria were between ages 18 and 65 yr with at least a high school education. Psychiatric or medical conditions, intellectual disability, developmental disorder, or atypical BSI scores were exclusion criteria. All the participants who completed the questionnaires were included.
Procedure
We recruited participants through social networks and personal communication using a snowball sampling method. After signing a consent form, 54 participants (33.75%) completed the questionnaires on a secure website and 106 (66.25%) completed them on paper. All received the SPQ and S–Hab–Q and provided background information. In addition, 54 participants (33.75%) received the BSI, and 96 participants (60%) completed the AASP.
Assessment Tools
Sensory Habituation Questionnaire
The final version of the S–Hab–Q (Ben-Sasson & Podoly, 2014, 2017) consisted of 25 items rated on a Likert scale ranging from 0 to 3. The total score was computed as a sum of items with a possible range of 0–75.
Sensory Perception Quotient Short Version
The SPQ Short Version is a 35-item self-report measure that assesses basic sensory function, that is, hypersensitivity (28 items) and hyposensitivity (7 items), across five modalities. Participants rated SPQ items on a Likert scale ranging from 0 (strongly agree) to 3 (strongly disagree). For easier readability of SPQ scores, items that identified hypersensitivity were reverse scored, so that a higher score indicated higher SOR. A SOR summary score was computed, and principal components analysis showed that most items loaded on one factor. In a previous study (Tavassoli et al., 2014), the SPQ construct validity was tested using SOR scales (Schoen et al., 2008) and found satisfactory (r = –.46). Another study (Ben-Sasson & Podoly, 2017) conducted with typical adults who completed the Hebrew SPQ Short Version indicated high internal reliability (α = .86). In the current sample, the internal reliability of the SPQ was also high (α = .82). The questionnaire was translated into Hebrew (Ben-Sasson & Podoly, 2014) with permission of the authors.
Adolescent/Adult Sensory Profile
In this 60-item self-report scale, designed to measure sensory processing in people ages 11–65 yr, each item describes a behavior related to an everyday sensory experience. The AASP generates four quadrants: sensory sensitivity, low registration, sensory avoidant, and sensory seeking. Higher scores indicate stronger expressions of the pattern. An AASP–SOR (low threshold) score was computed as a sum of sensitivity and avoidance. In the current sample, the internal reliability of the AASP was high (α = .87). The AASP was translated to Hebrew (Parush et al., 2006).
Brief Symptom Inventory
This self-report symptom inventory can be used as part of a psychological assessment. Distress levels during the prior week are rated for each of 53 symptoms using a 5-point Likert-type scale ranging from 0 = not at all to 4 = extremely. The inventory has nine symptom domains: Somatization, Obsessive–Compulsive, Interpersonal Sensitivity, Depression, Anxiety, Hostility, Phobic Anxiety, Paranoid Ideation, and Psychosis. In the current sample, the internal reliability of the BSI was excellent (α = .95). The BSI was translated to Hebrew (Canetti et al., 1994), and this study used Israeli BSI norms (M = .72, SD = .59; Gilbar & Ben-Zur, 2002).
Data Analysis
Descriptive statistics characterize the distribution of the sensory questionnaires’ scores. There were no missing data in the S–Hab–Q, SPQ, or AASP. To explore convergent and discriminant validity, we conducted Pearson’s correlations between the total SOR scores, SOR modality scores of the three instruments, and AASP seeking and low registration scores. Internal consistency was evaluated at the scale and item levels using Cronbach’s α coefficient. In addition, Spearman correlations between the sensory scores and the BSI subscales were conducted because the BSI subscales were not normally distributed. As a result of multiple comparisons within tests, we conducted Bonferroni corrections setting the α value threshold at .006.
Results
Sensory Habituation Questionnaire Descriptive Statistics
The mean S–Hab–Q total score was 23.66 (SD = 10.57, range = 4–60). Four items from the questionnaire received a particularly high score; that is, 40% of participants answered 2 (very long time) or 3 (I can’t get used to it). The item “If I tie my shoelaces more strongly on one foot, I will continue to feel the differences between my feet for some time” received the highest score (55.5% answered 2 or 3). In contrast, the item “I can’t eat for a while in the presence of sounds (TV, music) coming from the next room” received a low score (7.2% answered 2 or 3). This item was also the only one that participants answered uniformly: 118 (73.8%) responded 0 (not long at all).
Construct Validity
Convergent Validity
There was a moderate significant correlation between the S–Hab–Q and the SPQ total scores, r(160) = .57, p < .001, and between the S–Hab–Q total score and the AASP–SOR score, r(96) = .61, p < .001, consistent with H2 and H3. No significant difference was found between correlations (z = –.46, p = .31). Table 2 presents correlations between S–Hab–Q and SPQ modality scores.
Correlations Between the S–Hab–Q and the SPQ Modality Scores (N = 160)
Note. S–Hab–Q = Sensory Habituation Questionnaire; SPQ = Sensory Perception Quotient.
p ≤ .006.
Divergent Validity
The S–Hab–Q score correlated with AASP low registration, r(96) = .45, p < .001, but not with AASP seeking, r(96) = .09, p = .36, partially confirming H4. Table 3 presents correlations between the AASP–SOR modality scores and the other two perceptual measures, the S–Hab–Q and SPQ.
Correlations Between AASP–SOR and S–Hab–Q and SPQ Modality Scores (n = 96)
Note. AASP–SOR = Adolescent/Adult Sensory Profile–Sensory Overresponsivity; S–Hab–Q = Sensory Habituation Questionnaire; SPQ = Sensory Perception Quotient.
p ≤ .05. **p ≤ .006.
Table 4 presents Spearman correlations between sensory scores and the BSI. As hypothesized, the S–Hab–Q did not correlate with the BSI subscales, whereas most correlations between the AASP–SOR and BSI subscales were significant (p ≤ .006).
Correlations Between SOR Scores and BSI Subscales (n = 54)
Note. AASP–SOR = Adolescent/Adult Sensory Profile–Sensory Overresponsivity; BSI = Brief Symptom Inventory; S–Hab–Q = Sensory Habituation Questionnaire; SOR = sensory overresponsivity; SPQ = Sensory Perception Quotient.
p ≤ .006.
Internal Reliability
Internal reliability of the S–Hab–Q was high (α = .88), with no item reducing reliability, including items with low variability. Therefore, it confirmed H1.
Sensory Habituation Questionnaire and Demographic Correlates
Independent-samples t tests showed insignificant differences in S–Hab–Q total scores by gender (p = .620). A Spearman’s correlation between the S–Hab–Q total and age was weak, r(160) = –.24, p = .002. For education, independent-samples t tests indicated that S–Hab–Q total scores were significantly higher for people with nonacademic education (M = 25.9, SD = 11.72) than for those with academic (i.e., higher) education (M = 21.6, SD = 8.95), t(158) = 2.63, p = .009, d = 0.41.
Discussion
In this study, we describe the development of the S–Hab–Q, which aims to assess the slow habituation component of SOR. To date, sensory measures, including those evaluating SOR, have focused on sensitivity and negative affective reactions toward sensory stimuli. By mapping the difficulty-suppressing response in a timely manner, this questionnaire can explain the overload and distractibility associated with SOR. Our preliminary findings support the S–Hab–Q’s construct validity relative to other published sensory questionnaires and its independence from measuring psychopathological symptoms. A panel of experts approved the proposed questionnaire items, which ultimately yielded acceptable internal reliability.
Contrary to our hypothesis, the S–Hab–Q did not show a stronger correlation with the SPQ than the AASP–SOR. Moderate correlations were found among the S–Hab–Q, SPQ, and AASP–SOR. These correlations imply that the S–Hab–Q adds unique aspects not captured by the other tools. Correlations between comparable sensory modality scores were not particularly high relative to correlations across different modality scores, suggesting habituation should be treated as a unified construct rather than specific to a particular modality.
The SOR construct is separated from low registration and seeking in several measures (e.g., the AASP). We found a significant moderate correlation between the S–Hab–Q and the low-registration quadrant of the AASP. Surprisingly, slower rather than faster habituation was associated with high levels of low registration. Scholars have suggested that low registration reflects an attempt to shut down and avoid overstimulation (e.g., Lane, 2002). As such, low registration denotes for some people an underlying state of SOR. Interestingly, no correlation was found between the S–Hab–Q and the AASP seeking quadrant, in contrast to previous findings that linked rapid physiological habituation with a need for constant stimulation (Brown et al., 2001). Further physiological research and exploration of these correlations are warranted in future studies.
Results revealed that the S–Hab–Q and the SPQ do not measure levels of negative affect or psychological reaction to different stimuli, thus emphasizing their value as sensory–perceptual measures. This lack of correlation with emotional components contrasts with the AASP–SOR score, which did correlate with BSI scores, indicating the AASP–SOR’s divergence from the SPQ and the S–Hab–Q. The AASP–SOR score was previously associated with psychopathology such as depression (Serafini et al., 2016), anxiety (Levit-Binnun et al., 2013), and obsessive–compulsive disorder (Rieke & Anderson, 2009). It is difficult to tell which component of the AASP measurement—affective response, sensitivity, or habituation—contributes to these correlations. We must be cautious because the BSI is a screening tool; hence, there is a need to test the interplay of habituation and psychopathology using a diagnostic psychopathology tool.
Future research in clinical populations should enable the study of the accurate comorbidity rates between sensory abnormalities and psychopathology. The S–Hab–Q item distributions indicate at one extreme a very common ability to habituate quickly to auditory stimuli while eating but a relatively rare capacity of responders to habituate to their shoelace being tied asymmetrically. Although scale items were highly consistent, different stimuli yield different habituation rates. It is plausible that the typical quick habituation to noise while eating signifies the modulating effect of oral stimulation on auditory habituation (Williams & Shellenberger, 1996). This understanding corresponds with Ayres’ (1972) theory, which emphasized the importance of combining modalities to facilitate function. These item-level differences may also reflect differences in the commonality of the described experiences, with eating in a noisy environment representing a common experience and asymmetric shoe tying a rare one.
Investigating the relationship between S–Hab–Q scores and demographic information yielded no gender differences, in line with previous research (Tavassoli et al., 2014). However, in another Israeli sample using the AASP, females were significantly more sensitive than males (Engel-Yeger, 2012). This difference in findings may reflect the use of a measure that is more perception oriented (S–Hab–Q) than affective oriented (AASP).
We found a weak correlation between age and S–Hab–Q scores; younger people had slower habituation ability. Similarly, the AASP showed that older adults were less sensitive than young and middle-aged adults (Pohl et al., 2003). Our findings reinforce the use of the S–Hab–Q without the need for additional adjustments such as age cutoffs, given the modest relationships.
Limitations
Despite S–Hab–Q’s strengths, this study had drawbacks. Most significant were the sample’s lack of representation and small size. In addition, item analysis is needed to further revise items with poor variability and less discriminating power, and, to equally represent each modality as well as develop item sets for younger age groups, new items should be considered. Finally, there is a need to investigate additional psychometric features crucial for clinical implementation, such as discriminant validity in clinical populations and test–retest reliability. Although we interpret the lack of association between S–Hab–Q and BSI scores as reflecting the S–Hab–Q’s capacity to capture perception alone, it could be a result of the use of a neurotypical population and the restricted range of scores on the BSI.
Implications for Occupational Therapy Practice
The results of this study have the following implications for occupational therapy practice:
The S–Hab–Q shows potential to assess daily sensory habituation among adults.
Questionnaire items provide a wide view of habituation capacities and functional outcomes, facilitating more refined occupational therapy evaluation and intervention.
Together, the S–Hab–Q and the SPQ can be useful instruments to guide intervention that addresses both sensitivity and habituation aspects of SOR.
Differentiating sensitivity and habituation dimensions allows clinicians to psychoeducate clients regarding adjusting to aversive stimuli and encourages development of cognitive coping strategies when encountering them (e.g., “I am aware of this stimulus now, but I can still habituate. I just need more time”). As such. the S–Hab–Q has direct clinical relevance.
Conclusion
To date, habituation has been clinically overlooked but frequently reported by people with SOR. Introducing the S–Hab–Q is an important step in breaking down the SOR components while avoiding overlap with negative emotionality. Further research underway is investigating the correspondence of S–Hab–Q with unique electrophysiological patterns. Identifying slow habituation could have a positive impact on learning and daily functioning and can reduce vulnerability for those suffering from psychopathology.
Footnotes
Acknowledgments
This study was funded by the National Institute of Psychobiology in Israel.