In times of uncertainty, reality supersedes imagination. Yet when ‘reality strikes’, imagination maybe our only defence against stress and uncertainty. During pandemic, imagination can foster coping, well-being and recovery. Although imagination can cause pain and suffering for some (e.g., catastrophizing worst case scenarios); imagination can also foster change, comfort and hope under threat (Sheikh, 2003; Awad, 2017; Kalsched, 2021). Imagine a person sheltered-in-place using imagination to invent a new social platform, write a novel; or a six-year-old who pretends to be a ‘face-masked superhero’ saving the world from the next pandemic. Under continuous stress when reality becomes uncertain, both children and adults turn inward to imaginary worlds where they feel secure, gain control and clarity, play and envision solutions. Fantastic Reality (FR) is that imaginative sphere: “The link between the infinite ability of imagination to create a world, desired or required, and actions taken to solve problems that exist in reality” (Lahad, 2017a,b, p. 30). FR entails the ability to transcend to a fantastic realm where people feel impregnable and safe, and where they can manage, adjust and reconstruct the unbearable reality.

Both children and adults reframe meaning through imaginal coping, using spontaneous expressive forms of imagination, especially in times of stress and uncertainty. Such non-literal fantastic forms of expression stem from the human capacity for as if thinking, and provide affective relief and flexibility of meaning (i.e., play, humor, pretense, narrative, ritual, metaphor). Openness to flexible imaginings and playfulness as a means of coping can foster resilience (Clark, 2016). Over the lifespan, imagination can serve a fundamental role as a higher mental function in relation to coping and resilience (Awad, 2017). Nonetheless, further research is required to determine how and under what conditions imagination fosters resilience in response to stress and adversity (Capurso and Ragni, 2016; Rubinstein and Lahad, 2022).

Moreover, there is a paucity of quantitative research and tools measuring imagination in relation to stress and trauma. Valid and reliable instruments are needed to identify the determinants and predictors of imagination abilities (Jankowska and Karwowski, 2015; Jung et al., 2016). The FRAME was devised to clarify the role of imagination in coping with stress and trauma (Rubinstein et al., 2021). The COVID-19 provided an opportunity to measure FRAME responses in the general population coping with uncertainty and pandemic related stress.

Initial responses to the 21-item FRAME demonstrate high internal consistency (α = 0.88) and good test–retest reliability over 27 weeks (0.60 < r < 0.80; Rubinstein et al., 2021). Exploratory factor analyses (EFA) suggest that FRAME responses are composed of four distinct first-order factors, consistent with theory (Lahad, 2000, 2005; Lahad and Leykin, 2012; Rubinstein et al., 2021). These results are in accord with recent research supporting the assertion that imagination is not a singular construct, but instead composed of various interrelated elements (Liang and Chia, 2014; Zabelina and Condon, 2020).

FRAME factors are correlated in expected directions with ego resiliency, playfulness and fantasy proneness, suggesting convergent validity. Yet none are strongly correlated with divergent thinking, dissociation or PTSD symptoms, supporting the discriminant validity of responses (Rubinstein et al., 2021). FRAME playfulness, control, and coping factors are correlated with concepts associated with mental health and resilience (i.e., ego resiliency/playfulness) and psychopathology for the transcendence factor (i.e., dissociation, PTSD symptoms). Moreover, transcendence appears distinct from ego resiliency. Initial findings suggest the “dissociative” factor of transcendence contributes most to the associations between FRA and clinical measures. Questions arise from these associations; how is FRA concomitantly associated with both well-being and psychopathology? What are the adaptive/maladaptive abilities of imagination use during adversity or crises? Perhaps a more nuanced, factorial approach is required to delineate problematic from mental health-enhancing FRA.

This study was undertaken to further assess the psychometric properties of FRAME responses during the COVID-19 pandemic at a time of stress and uncertainty. Our intent was to examine how FRA may contribute to understanding of how people manage under adversity. This includes comparison with established measures of imagination and resiliency in relation to the validity of FRAME responses.

For this cross-sectional, population based study, we recruited Israeli adults using social media advertising as widely used in social science and mental health research (King et al., 2014). A two-week Facebook campaign was conducted July, 2020 for a total of 6,940 exposers. Data collection occurred after the first COVID-19 lockdown in Israel and before vaccines became available. Two lottery prizes of 500 shekels (₪NIS) were awarded to two randomly selected participants. Non-probability-quota sampling was applied so that the sample better reflected the demographic composition of the country (e.g., age, region). The sole inclusion criteria were 18+ years of age, complete FRAME responses and demographic information. This study received ethical approval from the Institutional Review Board (#7/2021–2) at Tel-Hai College, Kiryat Shemona, Israel.

Fantastic Reality Ability Measurement (FRAME; Rubinstein, 2020). The 21-item FRAME measures use of imagination in response to stress and trauma. Responses are provided on a Likert scale ranging from strongly disagree (1) to strongly agree (7). Initial research reports high internal consistency (α = 0.88) and good test–retest reliability over 26–28 weeks (0.60 < r < 0.80). Responses demonstrate concurrent validity in relation to ego resiliency, playfulness and fantasy proneness (Rubinstein et al., 2021).

Exploratory factor analyses (EFA) suggest FRAME responses measure four factors: Coping (5 items; e.g., “When I encounter difficulties or obstacles, I use my imagination to find alternatives to action and problem solving”); transcendence (6 items; e.g., “It happens that I find myself so involved in a fantasy or daydream that it feels as though it were really happening to me”); playfulness (6 items; e.g., “I enjoy taking part in social games”); and control (4 items; e.g., “I control my imagination and I can imagine anything I want”). Responses showed concurrent and factorial validity as playfulness, control, and coping were correlated with “resilient non-clinical” coefficients (i.e., ego resiliency/playfulness) and the transcendence factor with’clinical” measures’ (i.e., dissociation and PTSD symptoms; Rubinstein et al., 2021).

The original Connor-Davidson Resilience Scale (CDRISK; Connor and Davidson, 2003) is a 25-item measure of the ability to cope with adversity. Responses are provided on a Likert scale ranging from not true at all (0) to true nearly all the time (4). The brief CD-RISC-10 is a 10-item version of the CDRISK that measures hardiness, flexibility, sense of self-efficacy, ability to regulate emotion, optimism and cognitive focus/maintaining attention under stress (CDRISK-10; Campbell-Sills and Stein, 2007). The psychometric properties of CDRISK and CDRISK-10 responses have been broadly supported across settings and populations (Ye et al., 2017; Almeida et al., 2020), languages and cultures (Cheng et al., 2020; Chen et al., 2021; Wollny and Jacobs, 2021).

Four-Factor Imagination Scale (FFIS; Zabelina and Condon, 2020) is a 26-item instrument with responses reported on a Likert-type scale ranging from very inaccurate (1) to very accurate (6). The four factors measure frequency – the amount of time spent in imaginative states (e.g., “I find myself lost in imagination very frequently”), complexity – the detail with which one imagines (e.g., “My fantasies are less detailed than most people’s”), emotional valence – the degree to which one’s imagination tends to be positively or negatively valanced (e.g., “I visualize negative outcomes for the future of the world”), and directedness – a measure of how goal-oriented (vs. “free-floating”) one’s imagination tends to be (e.g., “My daydreams are directed towards a specific outcome”). Participants are asked to indicate how accurately each statement described them. FFIS responses show good internal consistency, α = 0.87 (0.75 < α < 0.94 across subscales), and convergent and discriminant validity relative to established scales measuring related and unrelated constructs, respectively (Zabelina and Condon, 2020; Zabelina et al., 2021).

Sociodemographic and COVID-19 Information. Participants were asked to provide descriptive information (e.g., age, gender, region). This included education and marital status. We also asked various questions regarding expose to COVID-19 (i.e., recovered), symptoms and impact on health and well-being (i.e., financial, social, mental and physical health).

Confirmatory factor analyses (CFA) were performed to replicate initial EFA findings suggesting that FRAME responses measure four first-order factors. We also set out to determine if these factors simply covary or whether they map onto a second or higher-order FRA latent construct. Three goodness-of-fit-indices are reported to assess the overall fit of CFA models: An incremental, an absolute, and a parsimonious fit index. The comparative fit index (CFI) is an incremental index representing the extent to which a hypothesized model is a better fit to data than the null model. Coefficients greater than 0.94 for the CFI indicate good model fit (Byrne, 2016). The standardized root mean square residual (SRMR) is an absolute index which represents the standardized difference between observed and predicted correlations within a hypothesized model. Finally, the root mean square error of approximation (RMSEA) is a parsimony index which represents the extent to which a hypothesized model fits data relative to the general population. Coefficients less than.055 for the SRMR and RMSEA suggest good model fit (O’Rourke and Hatcher, 2013).

Correlation coefficients were computed to demonstrate convergent and divergent validity of FRAME responses relative to established scales. SPSS version 26 was used to compute descriptive statistics and correlation coefficients. CFA was performed using AMOS 26.0.

We performed CFA to replicate and further examine the measurement properties of the 4-factor EFA model (Rubinstein, 2020). As hypothesized, each item significantly measured its respective factor, and each first-order factor significantly measures a second-or higher-order FRA latent construct. CFA was computed using maximum likelihood estimation (Byrne, 2016; see Figure 1).

After correcting for correlated error between 22 of 230 possible pairs, statistics indicated excellent goodness of fit between the higher-order model and data, χ² = 306.92, df = 163, p < 0.01. The comparative fit index is in ideal parameters (i.e., CFI > 0.94, CFI = 0.95), same too for the root mean square error of approximation (RMSEA <0.055, RMSEA = 0.045); the full 90% confidence interval for the RMSEA is in ideal parameters (0.037 < RMSEA CL₉₀ < 0.053), as is the standardized root mean residual (SRMR <0.055; SRMR = 0.047). Based on confidence intervals for the RMSEA statistic (Mac Callum et al., 1996), statistical power for this CFA model was estimated at 0.99 (n = 437, df = 178), sufficient to detect medium to small effects (see O’Rourke and Hatcher, 2013).

To examine convergent and discriminant validity, the FRAME scores were correlated with additional measures administered to participants. Table 3 shows correlations between the FRAME, its four factors, and measures of resiliency and imagination. In general, the FRAME showed small to strong associations with the measures of resiliency (CDRISK-10-Item General Measure). The highest resiliency correlations were with FRAME Playfulness and Control factors (r = 0.54 & r = 0.46, respectively). Self-Efficacy and Optimism resiliency factors presented most prominently with the FRAME and its subscales (0.29 < r < 0.51). FRAME scores (all factors) are moderately correlated with Resiliency (CDRISK-10), Complexity, Directedness and Frequency (FFIS; 0.31 < r < 0.44). Consistent with theory, these coefficients suggest concurrent validity of FRAME responses.

Moderate correlation between FRAME factors further suggests construct validity. For instance, Playfulness is strongly correlated with Resiliency (r = 0.54) and its subscales (0.32 < r < 0.51), and Complexity and Directedness of imagination (r = 0.40 & r = 0.31, respectively). Control is moderately correlated with Resiliency (r = 0.46) and its sub-scales (0.30 < r < 0.44), and with imagination Complexity (r = 0.36) and Emotional-Valence (r = 0.30). Coping is also moderately associated with Resiliency (r = 0.32). Finally, Transcendence is strongly associated with imagination Frequency (r = 0.65) and moderately associated with imagination Complexity (r = 0.32). Consistent with theory, playfulness, control and coping are correlated with ‘resilient’ coefficients. The transcendence factor was mostly correlated with measure of imaginations frequency (r = 0.65). These findings provide further support for the concurrent validity of the FRAME and its factors (see Table 3).

FRAME responses are not strongly correlated with the imagination Emotional-Valence (r = −0.19). Nor are FRAME responses significantly associated with sociodemographic variables such as age (r = −0.08, ns), gender (t [df = 435] = 0.47, ns), education (t [df = 435] = 1.35, ns) or region (t [df = 435] = 0.93, ns). The transcendence factor is unrelated to resilience measures.