The main idea of the PHT is to explore how casual events can inevitably play an important role in the development of career trajectories, and how events attributed to chance can often be indirect results of effective behavior (Mitchell et al., 1999). This theory identifies five skills that facilitate the recognition, creation, and maximization of the use of chance as a career opportunity: (a) curiosity: the ability to explore new learning opportunities, (b) persistence: the ability to continue their efforts in the face of challenges, despite setbacks, (c) flexibility: the ability to adapt, change or adjust their attitudes and circumstances in the face of uncertain contexts, (d) optimism: the ability to identify new opportunities and perceive them as possible and attainable, and (e) risk-taking: the ability to act in the face of uncertain results.

Considered one of the core skills identified by the PHT, career flexibility can be defined generically as the degree to which a decision leading to action allows the individual to have other choices over time (Yahanpath et al., 2013). According to this theory, one needs to be aware that unplanned events tend to occur, and having a flexible attitude makes it possible to be prepared to find, assimilate, and grasp those opportunities or events attributed to chance (Mitchell et al., 1999). Flexibility is essential, for example, to assess the risks associated with a particular course or career choice (Yahanpath et al., 2013). Together with optimism, it predicts psychological well-being and can be considered a significant individual strength that contributes to the sustainable development of each person in different contexts (Valickas et al., 2019).

August-Brady (2000) examined the concept of flexibility and identifies it as a complex and multidimensional construct defined as an evolutionary and resilient response to recognized changes and uncertainties based on openness and willingness to change. Flexibility allows attention to be directed towards a greater diversity of choices and greater effectiveness and efficiency in results; in this way, exploration tends to be a notable attribute of individuals with flexible attitudes (Kim et al., 2023).

Career flexibility can be characterized as having an active nature, that is, the ability to respond and adapt to different purposes, conditions, or changes when performing actions effectively and efficiently is evidenced. It is noteworthy that inherent to this active nature is the ability to change or respond to changes without external intervention. In other words, resilience is an attribute that seems to denote the active nature of flexibility (August-Brady, 2000).

However, Kim (2019) argues that, in the definition of career flexibility, only the active nature of the construct is considered (e.g., Mitchell et al., 1999; Krumboltz, 2009), when it would be ideal to add to the concept, a passive nature too. The passive nature of flexibility is characterized as a state that is easily changed by external factors and moves with pressure, that is, passively flexible individuals do not commit to actions toward a goal (August-Brady, 2000; Kim, 2019). The passive nature of flexibility encompasses states of stagnation and indecision, which, in turn, can play a dysfunctional role in career development and cause anxiety and psychological distress (Gati, 2013; Kim et al., 2020).

To assess career flexibility more accurately and considering its multidimensionality, that is, the passive nature and the active nature of the construct, Kim (2019) developed a new instrument, the Career Flexibility Inventory (CFI) and identified three main dimensions of career flexibility, namely: (a) wavering: represents the passive aspect of flexibility. Passively flexible individuals do not present attitudes and behaviors that prioritize their goals, as the behavior is moving by external stimuli and pressures, and not necessarily so that a goal is effectively achieved. Passively flexible people sometimes feel indecisive about career choices and goals, (b) active adaptation: represents the proactive aspect of flexibility and is characterized by active behavior in the face of opportunities, marked by a proactive approach to changes and that considers productive alternatives with certain long-term career goals, and (c) flexible thinking: characterized by the ability to accept unplanned changes and to present an open attitude toward possible instabilities in career trajectories. This factor represents the reactive aspect of career flexibility. In summary, wavering refers to the passive nature of flexibility, and active adaptation and flexible thinking represent the active nature of flexibility.

As a multidimensional construct, time perspective can be defined as the process by which the continuous flow of existence is grouped into three temporal categories, the past, the present, and the future. These categories help individuals to attribute order, coherence, and meaning to life events (Nuttin and Lens, 1985; Zimbardo and Boyd, 2008). Effectively, developing a balanced time perspective means having an awareness of what the past and present are and how to project themselves into the future, and this awareness is essential for individuals to be able to plan their goals and to make career decisions (Kirdök, 2018).

Individuals often exhibit a dominant time orientation, linked to various behavioral and psychological outcomes. For instance, individuals with a dominant future orientation focus predominantly on future goals and tend to be self-disciplined and perseverant. People with an orientation to the past tend to value traditions and have a posture of resistance to social changes, whereas individuals with a present orientation tend to be more impulsive and extroverted and are willing to focus on enjoying the moment (Zimbardo and Boyd, 2008).

Future orientation appears to be essential for motivation, engagement, and formation of career-related interests (Taber, 2013; Imbellone and Laghi, 2016), as well as for career maturity (Janeiro, 2010) and career decision-making self-efficacy (Park et al., 2020; Kvasková and Almenara, 2021). Low levels of future time perspective are associated with career choice anxiety (Park et al., 2018) and career indecision (Ferrari et al., 2010). Present hedonistic, a negative vision of the future and past negative orientation, all seem to contribute to various career decision difficulties (Taber, 2013) and, when taken together, also contribute to low career decision-making self-efficacy (Kvasková and Almenara, 2021).

In the educational context, academic performance and the scientific area of study have been objects of increasing attention in career psychology due to the remarkable impacts it can have on career decisions, course satisfaction, and future plans (e.g., Cox et al., 2016; Maksimovic et al., 2020; Bennett et al., 2021; Boo et al., 2022). When related to time perspective, studies indicate, for example, that future orientation has a positive relationship with higher academic levels (Ferrari et al., 2010) and students with this orientation tend to have a more goal-oriented behavior, which can result in better grades in the long run (Barnett et al., 2020). Present orientation, on the other hand, appears to be negatively associated with academic engagement. Regarding the scientific areas of study, Bennett et al. (2021) identified that medical and health students were more confident in their career decisions than colleagues who attended courses in other scientific areas. However, these same students were less aware of alternative career paths and less prepared to reorient their careers if necessary.

The main objective of this study is to adapt the CFI (Kim, 2019) to the Portuguese context. Additionally, we intend to explore the relationships between career flexibility and time perspective and to investigate whether these factors differ between student groups based on their academic grades and scientific area of study.

Given the multidimensionality of both constructs, we expect that their relationship will vary in different ways. Specifically, as active adaptation and flexible thinking represent the active nature of flexibility and future orientation refers to a positive view of the future with a focus predominantly on future goals, it is expected that:

The minimum sample size of this study was estimated based on the recommendations by Hair et al. (2018), who consider that for a significance level of 5%, and power of 80% it is necessary to include 5 to 20 observations per study variable. The requirements regarding the minimum sample size for the multivariate analysis of variance (MANOVA) (Hair et al., 2018) were also considered. Therefore, for this study, a priori minimum of 300 participants was estimated. Initially, 1.420 participants, randomly selected from different Portuguese higher education institutions, responded to the survey. However, 40 observations (2,82%) were considered null, due to non-responses. From the total sample of 1.380 participants, 423 are male students (31%) and 957 female students (69%), and the ages ranged between 18 and 34 years (M = 22.12; SD = 3.6).

The research sample included students from 262 different courses (bachelor’s, master’s, integrated master’s, doctoral, and other postgraduate courses). For data analysis, the courses were organized into seven scientific areas (based on FCT – Foundation for Science and Technology classification), so the sample comprises 44% students from Social Sciences, 21% of Engineering Sciences and Technology, 12% of Humanities, 11% of Medicine and Health Sciences, 5% of Agricultural Sciences, 4% of Natural Sciences, and 2% of Exact Sciences.

The academic grades were organized into five levels in accordance with the European rating comparability scale, namely: E – values between 10–11 (2.4% of students), D – values between 12–13, (22% of students), C – values between 14–15 (41.5% of students), B – values between 16–17, (23.8% of students), and A – values between 18–20, (4.3% of students).

The study was approved by the Deontology Committee of the School of Psychology of the University of Lisbon.

Some sample inclusion criteria were defined, namely: (a) higher education students and (b) students aged 18 to 34 years old, age group characterized as young adults (Levinson, 1986; Arnett, 2000). For sample selection, emails were sent to course coordinators asking them to invite their students to participate in the survey. This e-mail contained all the information related to the objectives of this study and the instructions for accessing the survey.

The instruments were presented online [Qualtrics Surveys platform (version 1.2020), 2005] and the link to the survey was active from January to March 2022; the response time for each participant was 15 min on average.

Participation was voluntary and was conditioned on acceptance of the Informed Consent Form, made available in an online format. Data collection respected all ethical requirements and met the criteria for participant anonymity.

(a) Career Flexibility Inventory (CFI; Kim, 2019) is an inventory organized by Kim (2019) to assess career flexibility. It consists of a total of 15 items distributed in three dimensions, each one with five items, namely: (a) wavering, comprising items such as “I tend to adjust my career to given circumstances without having any long-term outlooks”, (b) active adaptation, “When it comes to choosing my career, I try to find various alternatives”, and (c) flexible thinking, “It is okay for me to change my career.” Two reverse-worded items have been included in the inventory (items 14 and 15) to inhibit random responses. In the inventory, the respondent is asked to evaluate how well each item describes him or her on a 5-point Likert scale, from 1 (never true) to 5 (always true). The first validity studies of the inventory in a sample of South Korean college students indicated adequate psychometric properties, and the model hypothesized of three dimensions was also acceptable (CFI = 0.91; TLI = 0.90; RMSEA = 0.07) (Kim, 2019).

The translation of the English version of the inventory involved an evaluation panel composed of researchers with proficiency in English and Portuguese languages, knowledge of vocational psychology, and familiarity with the cultural context of this study. A translation from English into Portuguese was prepared with subsequent back-translation of the instructions and items (semantic and lexical meanings) (Hambleton, 2005). No significant discrepancies were found between the versions designed by the specialists, and, after the final analysis of the inventory, a definitive version of the CFI in Portuguese was approved. A pilot study was applied to five college students in Portugal. The participants were encouraged to expose their perceptions about the content expressed in each item and, in general, no relevant doubts or difficulties were indicated.

(b) Time Perspective Inventory (Short Form) (TPI-Short Form; Janeiro et al., 2017) is an instrument organized by Janeiro (2012) to assess time perspective. The short form of TPI (Janeiro et al., 2017), has 19 items distributed into four scales (a) Future orientation scale includes 6 items that assess positive attitudes toward the future (e.g., “I have lots of projects for my future”), (b) Present orientation scale has 6 items and assesses attitudes and beliefs about the present (e.g., “I think life should be lived day by day”), (c) Past orientation scale, with 3 items, assesses attitudes toward the past (e.g., “I would like to be a child again because everything was easier then”), and (d) Negative vision of the future scale (4 items) assesses negative or anxiety-laden perceptions about the future (e.g., “I am going into the future not by choice but because I cannot stop”). Responses are indicated on a seven-point Likert scale (1. Does not correspond at all to the way I think, to 7. Corresponds very strongly). A study with the Short Form TPI conducted with a sample of 11th-grade students indicated adequate psychometric properties. The four-factor solution showed a good fit, χ2(144) = 334,287, p < 0.001, χ2/df = 2.32, RMSEA = 0.06, CFI = 0.92, TLI = 0.91, SRMR = 0.07, with standardized factor loadings ranging from 0.46 to 0.90 (Janeiro et al., 2017).

TPI-Short Form reliability indices (α and ω) in this study’s sample were adequate for three scales, future orientation, present orientation, and negative future orientation. The past orientation scale obtained a less satisfactory coefficient (Table 1), very similar to those found in previous studies with secondary school students (Janeiro et al., 2017).

(c) Participants also completed a sociodemographic form to characterize the sample in terms of age, gender, course they were attending, and academic grades of the last year.

The normality of the data was verified using a criterion based on a sample larger than 300 participants (Kim, 2013), considering absolute values of skewness < 2 and kurtosis < 7 (West et al., 1996).

Due to the lack of empirical evidence on how the items of the CFI would aggregate in Portuguese samples, a two-step procedure was conducted. The first step involved exploring the aggregation of the items using an exploratory factor analysis (EFA) (Brown, 2006) with the principal axis factorization method and oblique Promax rotation. To analyze the data, the pattern matrix, and the structural matrix were considered (Field, 2013). The criterion for determining significant factor loadings was based on sample size. Specifically, and following Hair et al. (2018) guidelines for samples with at least 350 participants, factor loadings of 0.3 or greater were considered significant.

The criterion based on classical parallel analysis (Horn, 1965; Baglin, 2014) was used to define the number of factors to retain, and for that, the PC software package FACTOR was used (Lorenzo-Seva and Ferrando, 2021).

The second step aimed to confirm the aggregation structure of the items using confirmatory factor analysis (CFA). The statistics and quality-of-adjustment indices adopted were the values of the chi-square difference test on the Satorra-Bentler Scale, estimated by maximum likelihood (ML, a suitable estimator for data that violate the assumption of normality). In addition to these, criteria such as comparative fit index (CFI), goodness-of-fit index (GFI), root mean square error of approximation (RMSEA), and the Akaike information criterion (AIC) were considered (Marôco, 2021b). Additionally, the convergent validity was calculated using the average variance extracted (AVE) and Composite reliability (CR) (Hair et al., 2018).

It is noteworthy that the total sample of this study (N = 1.380) was randomly divided into groups comprising approximately 50% of the sample, with 713 participants included in the exploratory factor analysis and 676 participants in the confirmatory factor analysis.

The internal consistency reliability of the CFI was evaluated using Cronbach’s coefficient and McDonald’s Omega (McDonald, 1999; Cohen et al., 2014).

To explore the possible relationships between CFI and TPI, the results of a bivariate correlation with Pearson’s coefficient were analyzed (Callegari-Jacques, 2009; Field, 2013).

In addition, a Multivariate Analysis of Variance (One-way MANOVA) was performed to explore and evaluate possible differences in the factors ‘academic grades’ and ‘scientific areas of study’ on career flexibility and time perspective. As the assumption of multivariate normality tested by Mardia’s Test was not supported in this sample, the measure to test the statistical significance of group differences in the MANOVA was the Pillai’s criterion (Hair et al., 2018). The assumption of variance–covariance homogeneity in each group was evaluated with the Box M test, specifically: Bos M Test (M = 138,341; F (112, 65503.83) = 1,191; p = 0.083) for the ‘academic grades’ factor and Bos M Test (M = 194,886; F (168, 94970.52) = 1,116; p = 0.145) for the ‘scientific area of study’ factor.

When significant effects were detected by MANOVA, an ANOVA for each of the dependent variables was performed, followed by Tukey’s Post-hoc HSD test. The level of significance considered was α = 0.05 (Marôco, 2021a).

Statistical treatment of the data was performed using SPSS Statistic for Windows, v. 28 and SPSS Amos Statistic for Windows, v. 27 (IBM Corp., 2022).

Absolute values of asymmetry and kurtosis indicated the reasonable normal distribution of data. Using the Mahalanobis distance, ten observations were considered bivariate outliers. The mean, standard deviation, and range of the data were calculated with and without outliers and as no significant impacts were observed on the results the outliers were not removed (Manly and Navarro Alberto, 2017).

KMO results (0.80) and Bartlett test (p < 0.000) indicate that the EFA is adequate for the sample of this study. To determine the number of factors to retain it was used the parallel analysis procedure. This analysis indicated 3 factors to retain, explaining about 52.31% of the total variance.

Table 2 shows the results of the two matrices after oblique rotation and the factorial loadings of the CFI items.

The first factor had factor loadings greater than 0.5 in both matrices; it comprises the five items of the wavering dimension in the original inventory and explains 23.54% of the total variance. The second factor explains 19.52% of the variance and grouped seven items with factor loadings above 0.4, and one item with a loading value of 0.32. The highest factor loadings were assigned to five items that belonged to the flexible thinking dimension in the original inventory and two items from the active adaptation dimension. The third factor, which explains 9.26% of the total variance of the data, included only three items with loadings with values mostly greater than 0.6, which were part of the active adaptation dimension in the original version of the inventory.

Through CFA, we sought to evaluate the quality of fit of two hypothetical theoretical measurement models to the correlational structure of the CFI items. The first hypothetical model, called “H1:15-3 original version,” considered the structure of the CFI in terms of the grouping of the items in the original version. The second hypothetical model, called “H2:15-3PT,” assumed the results of the exploratory factor analysis performed with the Portuguese sample (Table 2). Both hypothetical models were restricted to first-order factors. Table 3 presents the summary of the goodness-of-fit indices of the hypothesized models.

As shown in Table 3, the model “H2:15-3PT,” had a poor fit (χ²/df [674,682/87] = 7.75; RMSEA = 0.100; CFI = 0.787; GFI = 0.879; AIC = 740.682). To improve it, was eliminated six outlier observations and we examined the modification indices (greater than 11; p < 0.001, Marôco, 2021b), respecting the theoretical considerations. The measurement errors of items 1, 2, 3, 4, and 5 of the wavering dimension and items 9, 10, 12, 13, 14, and 15 belonging to the flexible thinking dimension were correlated. In addition, correlations were considered between errors of items saturated in different factors, because, theoretically, according to Kim (2019), there are correlations between the dimensions. Specifically: correlations between errors in items 4 (wavering) and 6 and 7 (active adaptation), 1 (wavering) and 6 (active adaptation), 3 (wavering) and 15 (flexible thinking), 7 (active adaptation), and 13 (flexible thinking) and finally 8 (active adaptation) and 12 (flexible thinking) were assumed. In this way, it was possible to obtain a good fit quality (H2:15-3PT modified: χ²/df [262,379/68] = 3.86; RMSEA = 0.065; CFI = 0.931; GFI = 0.951; AIC = 366.739).

The “H2:15-3PT-modified” model (Table 3) presents a better fit compared to the other hypothetical models, including the model that supports the distribution of items considering the original version of the inventory (H1:15-3 original version).

Regarding the convergent validity, Table 4 shows the factor loadings, composite reliability, and average variance extracted.

As shown in Table 4, most of the factorial loadings presented values above 0.5, and only items 10 and 15 obtained lower values, both items of the Flexible Thinking dimension. The composite reliabilities of the three dimensions obtained satisfactory values, that is, >0.6 (Hair et al., 2018), with the highest value in Wavering (0.79). AVE estimates show less satisfactory indices of convergent validity in the three dimensions, Wavering had also the highest AVE value (0.43).

Table 1 presents the descriptive statistics of the three subscales of CFI (based on the structure of the CFA results for this Portuguese sample), the four subscales of TPI, as well as the correlations between subscales and the reliability coefficients.

Analysis of the correlations between the three dimensions of CFI (Table 1) revealed a positive and significant correlation between active adaptation and flexible thinking (r = 0.35, p < 0.01), by contrast, wavering established low correlations both with flexible thinking and with active adaptation, suggesting the relative independence between these dimensions of CFI.

Reliability indices show that two dimensions of CFI presented adequate values of internal consistency, namely wavering (α, ω = 0.78) and flexible thinking (α, ω = 0.76), however, the active adaptation dimension presented a lower value (α = 0.60, ω = 0.61).

The analysis of correlations between career flexibility and time perspective dimensions (Table 1) shows that future orientation had a moderate positive correlation with active adaptation (r = 0.33, p < 0.01) and a negative moderate correlation with wavering (r = −0.49, p < 0.01). The wavering subscale established a moderate and positive correlation with the negative future orientation (r = 0.52, p < 0.01).

To explore possible differences between academic grades and the scientific area of study on time perspective and career flexibility, a Multivariate Analysis of Variance (One-Way MANOVA) was performed.

According to the results of One-Way MANOVA, there is a significant difference in the academic grades of on time perspective variables and flexibility variables, Pillai’s Trace = 0.047, F (28, 5,160) = 2,18, p = 0.001, partial ƞ2 = 0.01, (π = 0.999). Given this, an ANOVA for each of the dependent variables was performed (Table 5).

The descriptive statistics in Table 5 show that among the dimensions of career flexibility, wavering was the dimension with the greatest variability across the five grade levels, while among time perspective dimensions the variability among groups seems larger. ANOVA results (Table 5) indicate that, in fact, there are, statistically significant differences between ‘academic grades’ in one of the career flexibility variables, wavering (F (4, 1,293) = 4.73, p = 0.001, partial ƞ2 =. 01) and in three variables of the time perspective, future orientation (F (4, 1,293) = 7.36, p = 0.000, partial ƞ2 = 0.02), present orientation (F (4, 1,293) = 3.39, p = 0.01, partial ƞ2 = 0.01) and negative future (F (4, 1,293) = 6.02, p = 0.000, partial ƞ2 = 0.02).

The results of Tukey’s post-hoc HSD test indicated significant differences in wavering between the groups of grade level E (M = 13.64; DP = 3.55) and level B (M = 11.59; DP = 3.85); in future orientation between level A (M = 26.85; SD = 8.53), B (M = 26.42; SD = 8.04) and level E students with the lowest average score (M = 22.42; DP = 7.18). In negative future between level E (M = 14.45; SD = 6.01), and B grade level (M = 10.35; SD = 5.51. In the dimension of present orientation, the differences were significant between grade level E (M = 25.42; DP = 5.65) and A and B grade level students (A: M = 21.14; SD = 7.29, B: M = 22.33; SD = 7.34).

According to the results of One-Way MANOVA, there is also a significant difference of the factor scientific area of study on time perspective and career flexibility variables, Pillai’s Trace = 0.073, F (42, 8,232) = 2,41 p = 0.001, partial ƞ2 = 0.012, (π) = 1.000. Then, an ANOVA was performed for each of the dependent variables.

ANOVAs results (Table 6), indicated a significant difference of the ‘scientific area of study’ on two variables of career flexibility, wavering (F (6, 1,373) = 4.34, p = 0.001, partial ƞ2 = 0.02) and flexible thinking (F (6, 1,373) = 2.99, p = 0.01, partial ƞ2 = 0.01), and two variables from the time perspective, future orientation (F (6, 1,373) = 2.91, p = 0.01, partial ƞ2 = 0.01) and negative future (F (6, 1,373) = 6.99, p = 0.01, partial ƞ2 = 0.03).

Tukey’s post-hoc HSD test shows statistically significant differences between the scientific areas in future orientation, namely between Medical and Health Sciences (M = 26.27; DP = 8.49) and Exact Sciences (M = 21.25; DP = 7.65). Significant differences in negative future appear between the scientific area of Medical and Health Sciences (M = 10.24; DP = 5.69), Social Sciences (M = 10.64; DP = 5.67) and Exact Sciences (M = 14.89; DP = 6.03).

Regarding career flexibility, there were differences in the wavering dimension between Exact Sciences, which had the higher score on this dimension (M = 14.71; DP = 4.27) and Agricultural Sciences (M = 11.44; SD = 3.63) and Medical and Health Sciences (M = 11.62; SD = 4.01), that had the lowest scores on this dimension.

Finally, in flexible thinking, statistically significant differences appeared between the groups of the scientific areas of Medical and Health Sciences (M = 24.76; DP = 4.57) and Engineering and Technology Sciences (M = 26.34; DP = 4.43).

By investigating the multidimensionality of career flexibility this study contributes to the theoretical and practical understanding of the nature of career flexibility and its role in career development. Even if in an indirect way, our findings highlight the need for vocational psychology to be aware of the unpredictability of professional trajectories, considering the ongoing social, economic, and educational transformations (Skrbiš and Laughland-Booÿ, 2019; Barbulescu et al., 2022). In this sense, developing skills that promote a positive and open attitude towards uncertainty can help higher education students manage career-related challenges and improve their psychological well-being (Kim et al., 2020; Gati and Kulcsár, 2021). These skills can be considered important personal resources that can help students navigate the unpredictable events of their careers and everyday life (Valickas et al., 2019).

Our findings also suggest that higher education institutions should be aware of the importance of these skills, promote their development and take actions to help their students feel more confident in their professional choices and committed to their plans. This includes considering events that may affect the educational and professional path and are attributed to chance, as highlighted in previous studies (e.g., Barbulescu et al., 2022; Kim et al., 2023).

There are several limitations to this study that should be acknowledged. A first limitation refers to the sample that cannot be considered representative of college students in Portugal, thus limiting the generalizability of the findings. Since the CFI is a relatively new measure, re-evaluating its factorial structure in different samples and contexts in Portugal could be worthwhile. Moreover, because this study is cross-sectional, it cannot establish causality between variables, future studies may include longitudinal research to expand the understanding of the nature of the relationships between these variables. Another set of limitations refers to the results of the study. In fact, results showed some weaknesses of the CFI, namely the modest variance explained by the three-factor structure, unsatisfactory results regarding convergent validity, and the less satisfactory reliability coefficient of the active adaptation dimension. These limitations underscore the importance of refining the measurement instrument and deepening the understanding of the nature of career flexibility. Therefore, future studies are needed to provide new information on the validation of the CFI. Subsequently, studies should explore, through structural equation modeling, the effects between these variables, as well as the potential mediation and moderation effects with other variables.

Furthermore, the relative scarcity of literature supporting the findings, especially since the CFI is based on relatively recent career studies, emphasizes the need for further research to expand the understanding of PHT.