Can Digital-Hoarding Behavior Exist Independently of Physical-Hoarding Behavior? A Latent Profile Analysis of Physical- and Digital-Hoarding Behaviors

Abstract

The first case of digital hoarding was reported in 2015. Research has largely used a variable-centered approach, which overlooks heterogeneity in hoarding behaviors. Moreover, whether digital hoarding can exist independently of physical hoarding remains controversial. This study employed a person-centered approach to address these gaps. We surveyed 551 social networking site users and conducted a latent profile analysis. Four profiles emerged: no-hoarding, mild-hoarding, digital-hoarding, and severe-hoarding. The digital-hoarding group showed the highest digital hoarding but the second lowest physical hoarding; the severe-hoarding group showed high levels of both. Digital-hoarding and severe-hoarding groups had significantly higher SNS addiction than the other two groups. The digital-hoarding group reported the highest job performance, life satisfaction, and happiness. These findings demonstrate that digital hoarding can exist independently of physical hoarding, carries both adaptive benefits (e.g., well-being, performance) and elevated SNS addiction risk, and calls for profile-based rather than one-size-fits-all interventions.

Plain Language Summary

Digital Hoarding Can Occur Independently But That Physical Hoarding Cooccurs Alongside Digital Hoarding Behavior

Digital hoarding refers to accumulating digital files (such as photos, documents, or videos) to the point where it becomes disorganized and stressful. Physical hoarding is a well known disorder involving excessive accumulation of physical items. In this study, we used a method that groups people based on similar patterns of behavior (called a person centered approach) and a statistical technique to identify natural groups (latent profile analysis). We surveyed 551 social media users and found four distinct groups: (1) people who do not hoard either digital or physical items; (2) people who hoard mildly in both areas; (3) people who mainly hoard digital items but not physical items; and (4) people who hoard severely in both areas. Importantly, the digital only hoarding group showed higher job performance, life satisfaction, happiness, and sleep quality than the other groups, but they also showed higher social media addiction. These findings suggest that digital hoarding can exist independently of physical hoarding and may have both adaptive benefits and potential risks. Our results can help design personalized interventions for different types of hoarding behavior.

Keywords

digital-hoarding behavior physical-hoarding behavior latent profile analysis person-centered approach SNS users

Introduction

The rapid development of the digital economy has resulted in a new type of behavior, namely, digital hoarding. The first case of digital-hoarding behavior was reported in 2015; furthermore, such behavior has been defined as “the accumulation of digital files to the point of loss of perspective, which eventually results in stress and disorganization” (van Bennekom et al., 2015). In addition, the implementation of physical isolation and lockdown measures as a result of the coronavirus disease 2019 (COVID-19) pandemic led to an increase in digital-hoarding behavior (Shauly et al., 2020). Digital-hoarding behavior is usually compared with physical-hoarding behavior, which is a more familiar topic. Within the context of physical-hoarding behavior, the estimated proportions of individuals who exhibit hoarding disorders range from 2% to 5% (Büscher et al., 2014) or from 1.5% to 5.8% (Cath et al., 2017) according to nationally representative population samples. Therefore, millions of people may engage in hoarding behavior. Moreover, hoarding behavior can occur in work environments (Sweeten et al., 2018). Importantly, digital- and physical-hoarding behaviors have serious medical and psychological ramifications for individuals who are exposed to these behaviors, including addiction, feelings of anxiety, sleep and health problems, obsessive compulsive disorder, and reduced cognitive performance (Grisham & Barlow, 2005; Neave et al., 2019, 2020; Sweeten et al., 2018; Thorpe et al., 2019; Woody et al., 2014). As the incidence of clinically relevant hoarding disorders increases by 3.7% with every year of a person’s age, appropriate treatments for both digital and physical hoarding are urgently needed (Cath et al., 2017). Therefore, exploring whether digital-hoarding behavior is related to or independent of physical-hoarding behavior offers great theoretical and practical value, particularly in the context of determining whether the treatments adopted for traditional physical-hoarding behavior are appropriate for digital-hoarding behavior.

Although we understand the consequences of hoarding behaviors, the question of whether digital hoarding can exist independently of physical hoarding remains controversial. To answer this question, we employed a person-centered approach to explore the profiles associated with hoarding behavior. The results of empirical studies in which a variable-centered approach has been used to investigate the relationship between digital and physical hoarding have been largely inconsistent (Howard & Hoffman, 2018). A variable-centered approach examines relationships among variables across all participants, assuming that the population is homogeneous; in contrast, a person-centered approach identifies subgroups of individuals who share similar response patterns, thereby capturing population heterogeneity (Gabriel et al., 2015; Peugh & Fan, 2013; Zhong et al., 2021). Accordingly, a new way of addressing this question is needed. Moreover, a variable-centered approach overlooks the possibility of heterogeneity in hoarding behaviors by assuming that the hoarding population is homogeneous. The inconsistency in the results that have been obtained via this variable-centered approach may be resolved by employing a person-centered approach to examine the possibility that different subgroups may participate in hoarding behaviors at quantitatively distinct levels (Gabriel et al., 2015; Zhong et al., 2021), thus highlighting the existence of different profiles of hoarding behaviors. Accordingly, in this study, we employed a specifically person-centered approach, which does not assume that the hoarding population is homogenous, to address the inconsistent findings that have been obtained via the variable-centered approach. To achieve this goal, we conducted a latent profile analysis (LPA) to group the participants in this study. The rationale underlying LPA focuses on model heterogeneity in cross-sectionally sampled data by grouping participants into latent categories on the basis of similarities in their response variable scores (Peugh & Fan, 2013, p. 617). Thus, this LPA approach could help us answer the question of whether digital-hoarding behavior is independent of physical-hoarding behavior.

Hypothesis Development

Theoretically, two perspectives on digital-hoarding behavior currently prevail. From the first and most common perspective, digital-hoarding behavior is regarded as a behavioral disorder that is similar to the well-known behavior of physical hoarding. This view logically extends to the claim that digital hoarding and physical hoarding share common psychological mechanisms and behavioral manifestations, which are particularly evident in individuals with severe psychological disorders. If digital- and physical-hoarding behaviors are essentially interrelated, as other researchers have proposed (S. Kim, 2013; Sweeten et al., 2018; Thorpe et al., 2019), these behaviors should be considered in similar ways. In practice, the sign and magnitude of the relationship between digital hoarding and physical hoarding are related directly to the corresponding treatment (Grisham & Barlow, 2005)

From the second perspective, digital-hoarding behavior is treated as a novel disorder that exhibits distinct psychological mechanisms and behavioral characteristics (van Bennekom et al., 2015). While digital-hoarding behavior may partly resemble hoarding or compulsive behaviors, it is nevertheless classified as a dysfunctional behavior (i.e., as a disorder). If digital hoarding is a completely unique disorder, as some researchers have suggested (Sedera & Lokuge, 2018; van Bennekom et al., 2015), we must develop a unique theory to explain its pathological and psychological mechanisms. For example, a thematic analysis revealed that physical and digital hoarding share similarities in terms of their focus on accumulation, difficulty discarding items and emotional distress (Sweeten et al., 2018). Furthermore, quantitative research has revealed that both digital and physical hoarding are related to obsessive compulsive disorder (Thorpe et al., 2019).

In contrast to these perspectives, the following novel theoretical hypothesis is proposed in this paper: Digital-hoarding behavior represents an adaptive and well-functioning strategy that can be used to manage digital information in the digital era when an individual faces the dilemma of infinite digital information and finite psychological resources. In the contemporary digital age, information explosion and digital overload have become ubiquitous (Sun & Lee, 2022), whereas individual cognitive resources (e.g., attention and memory) remain inherently limited. How can finite psychological resources be used to manage infinite digital information? Digital hoarding serves as an effective solution to this challenge (Sillence et al., 2026). Compared with physical hoarding, digital hoarding offers lower costs, simpler execution, higher efficiency, and more convenient retrieval. This behavior eliminates certain drawbacks of physical hoarding and may even offer unique advantages. Moreover, in terms of psychological mechanisms, Sweeten et al. (2018) identified the following key psychological mechanisms of digital hoarding: being emotionally attached to digital data, keeping digital data for the future/just in case, retaining digital data as evidence, and being reluctant to process digital data or files because of laziness or time constraints. Regardless of the specific psychological mechanism, a fundamental distinction from physical hoarding in the real world is that digital hoarding does not occupy the offline physical space of the user. In other words, the identified psychological mechanisms of digital hoarding may exist independently of the psychological mechanisms of physical hoarding because digital and physical hoarding belong to virtual-digital space and physical space, respectively. In light of this distinction, for a user with a normal life and work routine, physical space and online digital space can exist independently. Thus, in the digital era, many individuals may use digital technology for the purpose of hoarding, a novel normative behavior that is distinct from physical hoarding disorder. Therefore, we propose the following hypothesis:

Hypothesis 1: The digital-hoarding group represents a unique profile that is independent of the physical-hoarding group.

On the basis of our conceptualization of adaptive digital practices, we propose that the digital-hoarding group is associated with certain positive outcomes. However, we also recognize that digital hoarding may carry potential risks. In particular, it has been argued that hoarding is positively associated with addiction (Grisham et al., 2010). Thus, while digital hoarding may offer short-term adaptive benefits (e.g., psychological security, task efficiency), it may also predispose individuals to addictive social media use over the longer term. We therefore qualify our adaptive hypothesis by acknowledging this dual nature. Nevertheless, the potential for long-term risk does not logically preclude the possibility that digital hoarding, when occurring independently of severe physical hoarding, may still yield short-term adaptive advantages. In fact, precisely because digital hoarding lacks the spatial and material constraints of physical hoarding, it is plausible that individuals who engage primarily in digital hoarding (rather than severe combined hoarding) would experience better job performance, life satisfaction, happiness, and sleep quality. First, unlike individuals in the severe-hoarding group, those who engage in this form of adaptive digital-hoarding behavior can use digital tools to perform their jobs more efficiently (Neave et al., 2019; Sillence et al., 2026). As digitalization in the workplace advances rapidly and artificial intelligence (AI) tools are adopted widely, hoarded digital materials, such as documents, videos, and images, can exhibit synergy with digital work practices, thereby enhancing job performance. Notably, digital tools, such as those pertaining to communication and AI, are more helpful than severe-hoarding behavior is and can thus improve individuals’ job performance. Second, according to our conceptualization of adaptive practices, unlike excessive hoarding behavior, digital-hoarding behavior is a normal adaptive behavior. This adaptive digital behavior is consistent with the contemporary digital era and can help increase individuals’ life satisfaction and happiness (Valkenburg, 2022). For some individuals who engage in mild or moderate digital-hoarding behaviors, such behavior does not inherently encroach upon their physical reality, nor does it excessively disrupt their offline lives. Regardless of the extent of hoarding, it is essentially a matter of the user’s storage space. Thus, except in cases of severe digital hoarding, mild digital hoarding that is independent of physical hoarding may foster positive emotions, such as the psychological security of being able to access files at any time, potentially leading to higher levels of well-being. Third, unlike individuals in the severe-hoarding group, those in the normal digital-hoarding group exhibit functional emotional regulation ability and can manage their negative emotions; thus, they exhibit better sleep quality (Maranci et al., 2021). Accordingly, we propose the following hypothesis:

Hypothesis 2: Individuals in the digital-hoarding group exhibit higher levels of job performance, life satisfaction, happiness, and sleep quality than individuals in the severe-hoarding group do.

Methods

Participants and Procedures

The participant recruitment process for this research was influenced by both theoretical and practical considerations. At the theoretical level, our research question pertained to different types of hoarding behavior, that is, digital or physical; thus, we focused on users of online environments. At the practical level, these behaviors can be measured via online scales (Neave et al., 2019; Nutley et al., 2020). Moreover, as a result of the restrictions resulting from the COVID-19 prevention policies implemented during this study, online surveys represented an approach that was both appropriate and feasible. We adopted a convenience sampling method to recruit participants. The data referenced in this research were collected in May 2022. At the beginning of the questionnaire, the participants were informed of the purpose and procedure of this study as well as the confidentiality of their private information; importantly, they were informed that they could decide to cease participating in the study at any time. All of the scales used to measure the core variables were sourced from the extant, mature literature on this topic and have been reported to exhibit suitable validity and reliability (Wang et al., 2023). All of the scales were presented in Mandarin Chinese; scales that were originally developed in English were translated into Mandarin Chinese on the basis of Brislin’s (1980) translation–back translation procedure. A professor and six graduate students subsequently examined the semantics and relevance of the scales. Before the formal test was performed, a pilot study (n = 20) was conducted to refine the questionnaire and the administration procedure. The participants were given 5 ¥ (Chinese yuan) as compensation at the end of the study.

Ethics Statements and Declarations

We used the following methods and criteria to limit the risk of any potential harm to the participants. The participants in this research were at least 18 years old. Moreover, they were informed that they had the option to withdraw from the study at any point and that their information would remain anonymous; strict methods were used to protect their privacy. The participants could not be identified through this study. At the end of the study, no participants reported any harm. In contrast, the potential benefits of this research for understanding the potential treatment of hoarding disorders are significant. Thus, we believe that the potential benefits of this research to society outweigh the risk of harm to the study participants.

We obtained informed consent from all of the participants. Specifically, the participants were provided with full information regarding the purpose of and procedure used in this study, as well as other relevant information. After the participants read the content, they signed informed written consent forms. The questionnaire started with the declaration and purpose of the study. The study was conducted in accordance with the ethical standards of the authors’ institution and with the principles of the Declaration of Helsinki. Because it involved human participants, this study was reviewed and approved by the Ethics Committee of the author’s university.

Sample

Any participants who did not provide consent (5 participants) or who did not complete the survey (17 participants) were excluded from this research. No other participants were excluded. We recruited 551 participants; their demographic information is presented in Table 1.

Table 1.

Participant Demographic Characteristics (N = 551).

Variable	N (%)
Age (in years)
Mean (SD)	31.48 (7.24)
Gender
Male	226 (41%)
Female	325 (59%)
Education
<Elementary education	12 (2.2%)
High school	61 (11.1%)
Undergraduate	432 (78.4%)
Master’s	45 (8.2%)
Doctoral	1 (0.2%)
Annual family income (in 10,000 yuan)
<6	37 (6.7%)
6–9	56 (10.2%)
9–12	96 (17.4%)
12–18	123 (22.3%)
18–24	118 (21.4%)
24–36	77 (14%)
36–50	29 (5.3%)
>50	15 (2.7%)
Number of friends on WeChat
0–50	19 (3.4%)
51–100	58 (10.5%)
101–150	71 (12.9%)
151–200	84 (15.2%)
201–250	63 (11.4%)
251–300	72 (13.1%)
301–400	61 (11.1%)
401–500	56 (10.2%)
500–1,000	47 (8.5%)
1,000–2,000	13 (2.4%)
>2,000	7 (1.3%)
Work experience (including part-time work; in years)
1	58 (10.5%)
2	31 (5.6%)
3	31 (5.6%)
4	43 (7.8%)
5	58 (10.5%)
6	67 (12.2%)
7	48 (8.7%)
8	71 (12.9%)
>9	144 (26.1%)

Measures

Our research question and the extant literature on this topic inspired our choice of profile-defining variables as well as of the associated factors and outcome variables. First, we selected variables pertaining to digital-hoarding behavior (Neave et al., 2019) and physical-hoarding behavior (Nutley et al., 2020) because these factors are related to our research questions. Second, regarding the factors associated with the latent hoarding profiles, our variable selection was based on the characteristics of social media users that have been reported to be related to hoarding behavior. For example, research on digital-hoarding behavior among social media users has focused on social networking sites (SNSs), such as WeChat (Wang et al., 2023). Thus, we measured SNS addiction (Andreassen et al., 2012) and the amount of time spent on WeChat. Third, we collected data from workers. Accordingly, we measured variables pertaining to workers’ performance and well-being (Baer et al., 2015), including job performance, life satisfaction, job satisfaction, emotional exhaustion, happiness, likelihood of illness, sleep quality, and sleep quantity. We used these variables to explore the latent profiles pertaining to hoarding behavior as well as the associated factors and outcomes.

Physical-Hoarding Behavior

We measured physical-hoarding behavior on the basis of a 5-item online hoarding rating scale (Nutley et al., 2020). This scale has exhibited satisfactory and consistent reliability and validity with respect to the original physical hoarding scale (Nutley et al., 2020; Tolin et al., 2010, 2018). The participants were asked to indicate the extent to which they agreed with a series of statements on the following 9-point scale: 0 = “not difficult”; 2 = “mild difficulty; occasionally (less than once per week) acquire items that are not needed or acquire several unneeded items”; 4 = “moderate difficulty; regularly (once or twice per week) acquire items that are not needed or acquire unneeded items”; 6 = “severe difficulty; frequently (several times per week) acquire items that are not needed or acquire many unneeded items”; and 8 = “extreme difficulty; very often (daily) acquire items that are not needed or acquire many unneeded items.” A sample item is as follows: “Because of clutter or the number of possessions, how difficult is it for you to use the rooms in your home?” (Cronbach’s α = .84).

Digital-Hoarding Behavior

We measured digital-hoarding behavior via the digital-hoarding behavior scale (Wang et al., 2023; Wu et al., 2021). This behavioral scale was adapted from the digital-hoarding tendency scale (Neave et al., 2019) proposed by Wu et al. (2021) with the aim of developing a digital-hoarding behavior scale that exhibits good reliability and validity (Wu et al., 2021). We employed the 13-item Chinese version of the digital-hoarding behavior scale to measure the participants’ digital-hoarding behavior. The participants were provided with the following clear instructions: “We often use smartphones, computers, social media, cloud accounts, apps, USB drives, hard disks and other digital methods to store a wide range of digital materials, such as video or audio files, digital documents, digital forms, photos, music, email, software installation packages, game resources, web pages, and e-books. In the following items, the label ‘file’ refers to digital data in any form. Please carefully consider the files that you have on various digital storage devices before answering.” The participants were asked to indicate the extent to which they agreed with a series of statements on a 7-point scale (ranging from 1 = “strongly disagree”–7 = “strongly agree”). A sample item is as follows: “I accumulate files that others may not keep” (Cronbach’s α = .87).

COVID-19 Criticality

We measured COVID-19 criticality via a 3-item COVID-19 criticality scale (Lin et al., 2021) that was adapted from a previously developed criticality scale (Morgeson, 2005). The participants were asked to indicate the extent to which they agreed with a series of statements on a 5-point scale (ranging from 1 = “strongly disagree”–5 = “strongly agree”). A sample item is as follows: “I view coping with the COVID-19 pandemic as a high priority for attaining long-term success” (Cronbach’s α = .75).

Time Spent on WeChat

We assessed the amount of time that the participants spent on WeChat with the following question: “How many minutes did you spend on WeChat each day during the past week?” The participants were asked to respond to this question by selecting one of the following options: 1 = “less than 10 min,” 2 = “10–30 min,” 3 = “31–60 min,” 4 = “1–2 hr,”, 5 = “2–3 hr,”, 6 = “3–5 hr,”, 7 = “5–7 hr,”, and 8 = “7–10 hr,” or 9 = “more than 10 hr.”

Social Network Site (SNS) Addiction

We measured SNS addiction via a 6-item SNS addiction scale adapted from the Facebook addiction scale (Andreassen et al., 2012). The participants were asked to indicate the extent to which they agreed with a series of statements on a 5-point scale (ranging from 1 = “very rarely”–5 = “very often”). A sample item is as follows: “I feel an increasing urge to log onto social network sites (such as WeChat, Weibo, Facebook, or Twitter)” (Cronbach’s α = .75).

Job Performance

We measured job performance through the use of a 4-item scale (Baer et al., 2015) adapted from the job performance scale (MacKenzie et al., 1991). The participants were asked to indicate the extent to which they agreed with a series of statements on a 5-point scale (ranging from 1 = “strongly disagree”–5 = “strongly agree”). A sample item is as follows: “All things considered, I am outstanding at my job” (Cronbach’s α = .78).

Life Satisfaction

We measured life satisfaction with the following question: “How satisfied are you with your life?” The participants were asked to respond to this question on a 5-point scale (ranging from 1 = “very dissatisfied”–5 = “very satisfied”).

Job Satisfaction

We measured job satisfaction via a 3-item scale (Kreiner, 2006). The participants were asked to indicate the extent to which they agreed with a series of statements on a 5-point scale (ranging from 1 = “strongly disagree”–5 = “strongly agree”). A sample item is as follows: “Generally, I am very satisfied with my job” (Cronbach’s α = .73).

Emotional Exhaustion

We measured emotional exhaustion via a 3-item scale (Watkins et al., 2015). The participants were asked to indicate the extent to which they agreed with a series of statements on a 5-point scale (ranging from 1 = “strongly disagree”–5 = “strongly agree”). A sample item is as follows: “I feel emotionally drained after working” (Cronbach’s α = .83).

Happiness

We measured happiness with the following question: “How often were you happy during the past week?” The participants were asked to respond to this question on a 5-point scale (ranging from 1 = “almost never (fewer than one day)” to 5 = “always (seven days)”).

Likelihood of Illness

We measured the likelihood of illness with the following question: “How likely are you to get sick this year?” The participants were asked to respond to this question on a 5-point scale (ranging from 1 = “very unlikely”–5 = “very likely”).

Sleep Quality

We measured sleep quality with the following question: “How would you rate your sleep quality over the past week?” The participants were asked to respond to this question on a 5-point scale (ranging from 1 = “poor”–5 = “excellent”).

Sleep Quantity

We measured sleep quantity with the following question: “How many hours of sleep did you get last night?” The participants were asked to report their amount of sleep in hours.

Analytical Approach

An LPA was conducted to identify hoarding profiles (Gabriel et al., 2015) with the assistance of Mplus 8.3 software. This process began with a model in which two latent profiles were specified, and additional profiles were added until no further improvements in model fit were observed. To assess model fit, we examined the log likelihood (LL), Akaike information criterion (AIC), Bayesian information criterion (BIC), sample-size-adjusted BIC (SSA–BIC), Lo–Mendell–Rubin (LMR) likelihood ratio test, and entropy (Lo et al., 2001). The model that exhibited the best model fit would be associated with lower LL, AIC, BIC, and SSA–BIC values, higher entropy, and significant LMR results (p < .05), thus indicating that the specified solution featuring k profiles was significantly better than the solution featuring k–1 profiles.

In line with the suggestions of Lanza et al. (2013), we used R3STEP to assess the factors associated with the latent profiles. This approach involved a series of multinomial logistic regressions that were conducted with the assistance of Mplus to determine whether an increase in the number of factors would result in a corresponding increase in the probability of a given participant belonging to one latent profile group over another such group. In previous studies, the Bolck–Croon–Hagenaars (BCH) method has been employed to determine the consequences of latent profiles with the assistance of Mplus (Zhong et al., 2021). The BCH method involves conducting a weighted multiple group analysis to test the differences in outcomes between all profile pairs (Asparouhov & Muthén, 2014). This approach allowed us to examine whether certain outcomes were more common among individuals who were associated with certain profiles than among individuals who were associated with other profiles. A notable advantage of the automatic BCH approach is that it can be used to model classification uncertainty in latent profiles (Asparouhov & Muthén, 2014).

Results

Descriptive Statistics

The means, standard deviations, and correlations associated with our study variables are listed in Table 2.

Table 2.

Descriptive Statistics and Intercorrelations Among the Study Variables (N = 551).

Variables	M	SD	1	2	3	4	5	6	7	8	9	10	11	12
1. Digital hoarding	4.91	0.93	−
2. Physical hoarding	4.21	1.57	.28^**	−
3. COVID-19 criticality	3.59	0.82	.28^**	.22^**	−
4. Time spent on WeChat	4.62	1.58	.19^**	.19^**	.13^**	−
5. SNS addiction	3.33	0.73	.45^**	.38^**	.35^**	.18^**	−
6. Job performance	3.60	0.70	.32^**	−.03	.20^**	.01	.12^**	−
7. Life satisfaction	3.75	0.78	.12^**	−.05	.09^*	.002	.002	51^**	−
8. Job satisfaction	3.72	0.77	.16^**	−.15^**	.09^*	−.06	.003	.60^**	.60^**	−
9. Emotional exhaustion	2.48	0.92	.04	.35^**	.19^**	.13^**	.26^**	−.28^**	−.36^**	−.56^**	−
10. Happiness	2.69	0.66	.12^**	−.14	.09^*	.11^*	.02	.38^**	.52^**	.42^**	−.32^**	−
11. Likelihood of illness	2.42	0.86	.07	.17^**	.03	.02	.09^*	−.19^**	−.23^**	−.28^**	.25^**	−.22^**	−
12. Sleep quality	3.77	1.00	.08	−.09^*	.06	−.003	.004	.30^**	.42^**	.36^**	−.27^**	.36^**	−.19^**	−
13. Sleep quantity	7.19	1.10	.03	−.04	.04	.03	−.03	.20^**	.27^**	.24^**	−.21^**	.17^**	−.21^**	.48^**

p ≤ .05. ^**p ≤ .01.

Confirmatory Factor Analysis

A confirmatory factor analysis (CFA) was conducted to determine whether our focal variables were valid. Before we performed this LPA, we first conducted a CFA to investigate the focal variables. As indicated in Table 3, our hypothesized 2-factor model exhibited acceptable model fit, thus suggesting that our focal variables exhibited discriminant validity.

Table 3.

CFA Results.

Model	χ²/df	CFI	TLI	RMSEA	SRMR	Δχ²
Model 1: Hypothesized 2-factor model	2.50^***	0.94	0.93	0.05	0.05
Model 2: Digital and physical hoarding combined into a single factor	6.71^***	0.76	0.72	0.10	0.10	557.21^***

Note. N = 551. CFI = comparative fit index; TLI = Tucker–Lewis index; RMSEA = root mean square error of approximation; SRMR = standardized root mean squared residual.

***

p < .001.

Identification of Hoarding Profiles

According to the literature (Gabriel et al., 2015; Zhong et al., 2021), the best model should exhibit the following fit statistics: lower LL, AIC, BC, and SSA–BIC values and higher entropy. Additionally, the LMR results should be significant (p < .05). The fit statistics of the hoarding disorder profile structure developed for this study are detailed in Table 4. We selected the 4-profile solution because it yielded lower LL, AIC, BC, and SSA–BIC values than the other profile solutions did. Moreover, the 4-profile solution exhibited the highest entropy values and significant LMR results. Furthermore, the 4-profile solution ensured the parsimony and meaningfulness of the retained profiles. Overall, these theoretical and statistical considerations suggested that the 4-profile solution was the best model with respect to our data.

Table 4.

Fit Statistics for the Solutions Involving Different Numbers of Profiles.

Number of profiles	LL	FP	AIC	BIC	SSA–BIC	Entropy	LMR (p)
2	−17,825.350	55	35,760.699	35,997.845	35,823.251	0.894	.0000
3	−17,484.774	74	35,117.549	35,436.617	35,201.709	0.889	.0007
4	−17,251.330	93	34,688.660	35,089.651	34,794.429	0.890	.0026
5	−17,117.072	112	34,458.144	34,941.059	34,585.522	0.878	.7301
6	−17,021.746	131	34,305.491	34,870.328	34,454.478	0.889	.2500
7	−16,965.861	150	34,231.722	34,878.482	34,402.317	0.885	.6887
8	−16,898.288	169	34,134.576	34,863.259	34,326.781	0.886	.6701

Note. N = 551. LL = log likelihood; FP = free parameters; AIC = Akaike information criterion; BIC = Bayesian information criterion; SSA–BIC = sample-size-adjusted BIC; LMR = Lo–Mendell–Rubin likelihood ratio test.

Descriptive information regarding the retained profiles is provided in Table 5. We labeled the set of individuals who exhibited the lowest levels of digital- and physical-hoarding behavior the no-hoarding group (16.15%; M_{digital hoarding} = 3.41; M_{physical hoarding} = 3.13), as these individuals’ behaviors were very infrequent. The most common profile, that is, the mild-hoarding group (34.12%; M_{digital hoarding} = 4.62; M_{physical hoarding} = 4.08), consisted of individuals who reported low levels of both digital- and physical-hoarding behaviors. As illustrated in Figure 1, individuals in the digital-hoarding group engaged in the highest levels of digital hoarding and the second lowest levels of physical hoarding (26.32%; M_{digital hoarding} = 5.63; M_{physical hoarding} = 3.23). Finally, individuals in the severe-hoarding group exhibited high levels of both digital- and physical-hoarding behaviors (23.41%; M_{digital hoarding} = 5.57; M_{physical hoarding} = 6.29). Taken together, these results revealed one qualitatively different profile (i.e., the digital-hoarding group) and three quantitatively different profiles (i.e., the no-hoarding group, mild-hoarding group, and severe-hoarding group), thus suggesting that hoarding behaviors are characterized by both quantitative and qualitative heterogeneity. Importantly, these results also suggest that digital hoarding can occur independently but also that physical hoarding co-occurs with digital hoarding. Thus, Hypothesis 1 was supported.

Table 5.

Characteristics of the Retained Profiles.

Profile	N, % of sample	Digital-hoarding behavior	Physical-hoarding behavior
1) Mild-hoarding group	188, 34.12%	4.62 (0.40)	4.08 (1.06)
2) No-hoarding group	89, 16.15%	3.41 (0.58)	3.13 (1.13)
3) Digital-hoarding group	145, 26.32%	5.63 (0.45)	3.23 (0.89)
4) Severe-hoarding group	129, 23.41%	5.57 (0.56)	6.29 (0.87)

Note. N = 551. The values presented in the first column indicate the existence of the four retained profiles. The values presented in the second column indicate the corresponding number and percentage of the participants in the identified group. The values presented in the third and fourth columns indicate the means (standard errors) of the profiles.

Figure 1.

Latent profiles pertaining to the different hoarding behaviors (the y-axis indicates the participants’ mean levels of hoarding behavior). Individuals in the no-hoarding group did not engage in digital or physical hoarding; individuals in the mild group engaged in low levels of both digital and physical hoarding; individuals in the digital-hoarding group engaged in the highest levels of digital hoarding and the second lowest levels of physical hoarding; and individuals in the severe-hoarding group engaged in high levels of both digital and physical hoarding.

Factors Associated With Hoarding Behavior

As suggested, the R3STEP command in Mplus combines a series of multinomial logistic regressions and offers various advantages over traditional regression analyses (Gabriel et al., 2015; Zhong et al., 2021). Thus, we used the R3STEP command to model the factors associated with hoarding behavior. As indicated in Table 6, our results revealed that individuals who exhibited COVID-19 criticality were more likely to be assigned to the digital- or severe-hoarding group than to the mild- or no-hoarding group. Moreover, greater amounts of time spent on WeChat were more likely to be associated with assignment to the severe-hoarding group than with assignment to the no- or the digital-hoarding group.

Table 6.

Factors Associated With the Various Hoarding Profiles (R3STEP Results).

Factors	Mild-hoarding vs. no-hoarding group	Mild-hoarding vs. digital-hoarding group	Mild-hoarding vs. severe-hoarding group	No-hoarding vs. digital-hoarding group	No-hoarding vs. severe-hoarding group	Digital-hoarding vs. severe-hoarding group
COVID-19 criticality	0.199	−0.545^**	−0.908^***	−0.744^***	−1.107^***	−0.364
Time spent on WeChat	0.064	0.137	−0.141	0.073	−0.206^*	−0.279^**

Note. N = 551. All of the analyses were conducted via the R3STEP command in Mplus. R3STEP automatically conducts a series of multinomial logistic regressions to test whether an increase in a factor would result in a higher probability of an individual belonging to one group over another group. Positive (negative) values indicate that higher values for the antecedents render a participant more likely to be assigned to the first (second) of the two latent profiles included in the comparison. For example, the last number (−0.279) indicates that compared with users in the digital-hoarding group, users in the severe-hoarding group are more likely to spend time on WeChat.

p ≤ .05. ^**p ≤ .01. ^***p ≤ .001.

Hoarding Behavior Outcomes

In line with the suggestions of previous LPA studies, the BCH method was used to model the outcomes of the identified profiles given that this approach offers several advantages over traditional regression methods (Zhong et al., 2021). Thus, we used the BCH method to test the outcomes of hoarding behavior. As indicated in Table 7, our results revealed that the participants in the digital- and severe-hoarding groups exhibited higher levels of SNS addiction (M_{digital hoarding group} = 3.42, M_{severe hoarding group} = 3.78) and better job performance (M_{digital hoarding group} = 3.89, M_{severe hoarding group} = 3.74) than did those in the mild- or no-hoarding group (M_{mild hoarding group} = 3.24, M_{no hoarding group} = 3.37; M_{mild hoarding group} = 2.73, M_{no hoarding group} = 3.39; ps < .05). In terms of satisfaction, the participants in the digital-hoarding group exhibited the highest levels of life satisfaction (M_{digital hoarding group} = 3.90, M_{mild hoarding group} = 3.63), while those in the severe-hoarding group exhibited the highest levels of job satisfaction (M_{severe hoarding group} = 3.48 M_{mild hoarding group} = 3.20); the values for both of these groups were significantly greater than those observed among the participants in the mild-hoarding group (p < .05). In terms of emotional exhaustion, the participants in the severe-hoarding group (M_{severe hoarding group} = 2.86) exhibited higher levels of exhaustion than did those in the other three groups (M_{mild hoarding}_group = 2.52, M_{no hoarding group} = 2.23, M_{digital hoarding group} = 2.24; ps < 0.05). Furthermore, the participants in the digital-hoarding group (M_{digital hoarding group} = 2.86) exhibited higher levels of happiness than did those in the other three groups (M_{mild hoarding group} = 2.63, M_{no hoarding group} = 2.68, M_{severe hoarding group} = 2.59; ps < .05). In terms of the likelihood of illness, the participants in the severe-hoarding group (M_{severe hoarding group} = 2.61) reported higher levels of illness than did those in the no- and mild-hoarding groups (M_{mild hoarding group} = 2.39, M_{no hoarding group} = 2.27; ps < .05). In terms of sleep, the participants in the digital-hoarding group (M_{digital hoarding group} = 4.00) exhibited higher levels of sleep quality than did those in the mild-hoarding group and the no-hoarding group (M_{mild hoarding group} = 3.66, M_{severe hoarding group} = 3.70; ps < .05); however, no significant differences in sleep quantity were observed among the four groups. In conclusion, the participants in the digital-hoarding group reported the highest levels of job performance, life satisfaction, happiness, and sleep quality. Thus, Hypothesis 2 was supported.

Table 7.

Hoarding Profile-Associated Outcomes (According to the BCH Method).

Outcome	Mild-hoarding (A)	No-hoarding (B)	Digital-hoarding (C)	Severe-hoarding (D)	Chi-square value of the overall test	p Value
SNS addiction	3.24_{B, C, D}	2.74_{A, C, D}	3.42_{A, B,D}	3.78_{A, B, C}	122.98	.00
Job performance	3.37_{C, D}	3.39_{C, D}	3.89_{A, B}	3.74_{A, B}	54.22	.00
Life satisfaction	3.63_C	3.70	3.90_A	3.79	8.52	.04
Job satisfaction	3.20_{C, D}	3.19_{C, D}	3.31_{A, B,D}	3.48_{A, B, C}	32.66	.00
Emotional exhaustion	2.52_{B, C, D}	2.23_{A, D}	2.24_{A, D}	2.86_{A, B, C}	35.56	.00
Happiness	2.63_C	2.68_A	2.86_{A, B,D}	2.59_C	11.20	.01
Likelihood of illness	2.39_D	2.27_D	2.37	2.61_{A, B}	7.72	.05
Sleep quality	3.66_C	3.72	4.00_{A, D}	3.70_C	9.21	.03
Sleep quantity	7.10	7.24	7.29	7.19	1.96	.58

Note. N = 551. All of the analyses were conducted via the BCH method in Mplus. BCH analysis is used to conduct weighted multiple group analysis and test whether one group significantly differs from another group while avoiding group membership shifts. The values pertaining to SNS addiction, job performance, life satisfaction, job satisfaction, emotional exhaustion, and happiness represent the means for each profile. The subscripts indicate that the profiles exhibit significant differences at p < .05. For example, the first number, 3.24, is the outcome mean (in this case, the SNS addiction mean) for the mild-hoarding group. The subscripts indicate that in terms of SNS addiction, the mild-hoarding group is significantly different from the no-, digital-, and severe-hoarding groups. Specifically, the mild-hoarding group exhibited a significantly higher mean score for SNS addiction than the no-hoarding group did, but this mean score was significantly lower than those associated with both the digital-hoarding group and the severe-hoarding group.

Discussion

In this research, we employed a person-centered approach to explore various types of hoarding behavior. We identified four hoarding behavior profiles: a no-hoarding group, mild-hoarding group, digital-hoarding group, and severe-hoarding group. The participants in the mild-hoarding group reported low levels of physical and digital-hoarding behavior. The participants in the digital-hoarding group reported high (low) levels of digital (physical) hoarding behavior. Finally, the participants in the severe-hoarding group reported high levels of both digital- and physical-hoarding behaviors. The two hypotheses proposed in this study were thus supported: digital hoarding exists independently of physical hoarding, and digital hoarding is associated with positive outcomes.

Theoretical Contributions

Our study makes several key theoretical contributions to the digital-hoarding literature. First, we found that digital hoarding may exist independently of physical hoarding, suggesting it can reflect short-term adaptive practices rather than maladaptive hoarding. By identifying four distinct profiles varying in digital- and physical-hoarding behaviors, we demonstrate that physical and digital hoarding can either co-occur or occur independently. This heterogeneity supports different theoretical perspectives (S. Kim, 2013; Sedera & Lokuge, 2018; Thorpe et al., 2019; van Bennekom et al., 2015) within distinct subgroups. Although physical and digital hoarding share similarities—particularly distress (Sweeten et al., 2018; Thorpe et al., 2019)—our research supports both co-occurrence and independence, offering a novel perspective to resolve debates about whether digital hoarding is functionally identical to physical hoarding.

Notably, the Digital-hoarding profile demonstrated higher SNS addiction than No- and Mild-hoarding groups, alongside adaptive benefits (heightened job performance, life satisfaction, happiness). These benefits may reflect short-term advantages of maintaining extensive digital collections: retaining digital content reduces anxiety about information loss, supports memory retrieval, and facilitates task completion. However, elevated SNS addiction scores may signal longer-term maladaptive consequences. Over time, accumulating digital possessions—particularly interactive and socially rewarding social media content—may foster compulsive checking, fear of missing out, and excessive reassurance-seeking (Brand et al., 2019). What begins as functional preservation may evolve into compulsive over-engagement meeting behavioral addiction criteria.

One interpretation is functional co-occurrence: elevated SNS engagement may facilitate accumulation and use of digital artifacts that support productivity, memory, and social connectedness, yielding short-term benefits. An alternative, complementary explanation is a temporal trade-off: digital hoarding may confer immediate adaptive advantages but, for some individuals, evolve into compulsive SNS-driven patterns producing long-term maladaptive consequences (Sweeten et al., 2018). Individuals who hoard digital content extensively may spend more time on SNS platforms, increasing exposure to addictive patterns and reinforcing compulsive behaviors through intermittent social reinforcement. This highlights digital hoarding’s dual nature: adaptive intentions (maintaining valued content, curating social connections) may coexist with maladaptive processes related to compulsive social media engagement. Thus, digital hoarding may represent an initially adaptive strategy that, if unregulated, contributes to problematic SNS use through fear of losing digital possessions, difficulty discarding socially relevant content, and progressive blurring of instrumental use into compulsive overuse (van Bennekom et al., 2015).

The distinction between Digital-hoarding and Severe-hoarding profiles is instructive here. The Severe-hoarding group, exhibiting high levels of both digital and physical hoarding, reported lower well-being, suggesting a potential progression. The Digital-hoarding profile may represent an early-stage or domain-specific manifestation that still confers benefits. Without intervention, these behaviors may generalize to the physical domain and become increasingly disorganized, culminating in the pervasive impairments characteristic of the Severe-hoarding profile (Thorpe et al., 2019; van Bennekom et al., 2015). Thus, digital hoarding’s consequences appear time-dependent: short-term functioning and satisfaction may coexist with emerging longer-term risks for addiction and cross-domain hoarding.

Second, our findings challenge characterizations of digital hoarding as inherently pathological (Sweeten et al., 2018; van Bennekom et al., 2015). For the mild-hoarding group—the most common profile (34.12% of the sample)—digital hoarding appeared to serve as a normal coping mechanism for managing digital overload, associated with greater happiness, better sleep, and higher performance. These findings suggest reconsidering whether hoarding behaviors used to manage digital demands necessarily indicate disorder or direct links to physical hoarding (Thorpe et al., 2019). For most individuals, normative levels of digital hoarding may represent typical digital behavior rather than pathology.

Third, different hoarding profiles were associated with distinct factors. COVID-19 criticality was associated with digital- and severe-hoarding profiles, potentially reflecting psychological insecurity and hoarding as coping during the pandemic (Charilaou & Vijaykumar, 2023; Zhao et al., 2022). Time spent on WeChat was associated with severe hoarding, consistent with social media addiction literature (Nur-A Yazdani et al., 2022; Whelan et al., 2020). Regarding outcomes, the severe-hoarding group exhibited greatest SNS addiction, highest emotional exhaustion, and highest illness likelihood—aligning with research on hoarding’s negative consequences (Sweeten et al., 2018; van Bennekom et al., 2015). In contrast, the digital-hoarding group exhibited highest job performance, life satisfaction, happiness, and sleep quality. Although counterintuitive given prevailing views, this may reflect alignment with pandemic-driven transitions to online work and social interaction, where digital-hoarding behaviors matched job demands and lifestyle adaptations.

Practical Implications

Our results carry several practical implications. First, policymakers and medical personnel should prioritize identifying individuals exhibiting the severe-hoarding profile. Given their high levels of both digital and physical hoarding, these individuals are likely to experience reduced satisfaction, job performance, and sleep quality, alongside increased SNS addiction—warranting special attention.

Second, digital hoarding can occur independently of physical hoarding. Currently, understanding of its antecedents and consequences remains limited; further research is needed to determine appropriate treatment approaches (van Bennekom et al., 2015). Our study provides initial evidence that, for some individuals, digital-hoarding behavior represents an independent type warranting unique treatment considerations. Nevertheless, scale assessment alone is insufficient for diagnosing hoarding disorder, which typically requires multiple methods—including scales, clinical interviews, and other investigations. Hoarding disorder should therefore be diagnosed with considerable care.

Third, hoarding behavior is characterized by substantial heterogeneity, requiring personalized intervention strategies that account for different hoarding types. Notably, our findings indicate that digital-hoarding behavior can occur independently of physical-hoarding behavior, highlighting the need to move beyond intervention approaches that view digital hoarding solely through a disorder-based lens (Thorpe et al., 2019). These findings have implications for clinicians, digital health practitioners, and workplace managers. For clinicians, the study highlights the need to reconsider practices that treat digital hoarding categorically as a hoarding disorder (van Bennekom et al., 2015). For digital health practitioners, recognizing that not all digital-hoarding traits are as harmful as physical hoarding is essential, requiring case-specific analyses of advantages and disadvantages rather than applying attitudes from physical hoarding disorder literature (Nutley et al., 2020). For workplace managers, digital-hoarding behavior may actually enhance job performance and well-being in digital work environments for some individuals. Recognizing digital hoarding’s independence from physical hoarding—and reframing it within virtual space—is essential. Interventions designed for physical hoarding require evidence-based research to determine suitability for digital hoarding.

Finally, while the Digital-hoarding profile exhibited higher performance and well-being, it also showed greater SNS-addiction risk relative to No- and Mild-hoarding profiles. Interventions should not pathologize digital hoarding outright, which could disrupt genuine short-term benefits. Instead, interventions should help individuals cultivate sustainable digital habits—retaining organizational and memory-keeping strategies that support performance while introducing safeguards against compulsive over-accumulation and SNS engagement that may lead to long-term harm.

Limitations and Directions for Future Research

This study has several limitations. First, all participants were experienced internet users. Although advantageous for examining digital behaviors, focusing on SNS users may introduce sample bias; non-internet users may exhibit different profile structures. Additionally, cultural factors may influence digital- and physical-hoarding behaviors. Users from collectivist cultural backgrounds may be more susceptible to interpersonal influences, potentially increasing needs for offline social interactions to fulfill emotional needs related to social worth and belonging (Mesquita, 2001). Consequently, the interdependent relationship between digital and physical hoarding might be more pronounced in such contexts. Moreover, social media users from collectivist cultures may be more likely to hoard digital files for future use, suggesting a “prepared” mentality shaping perceptions of digital accumulation. For instance, individuals from collectivist cultures place greater weight on the future during delay discounting tasks (B. Kim et al., 2012). Future research should investigate how cultural orientations influence perceptions of and engagement in digital-hoarding behavior.

Second, digital hoarding is not universally adaptive, but that its consequences are time-dependent. The short-term profile of a digital hoarder may be one of high functioning and satisfaction, even as longer-term risks for addiction and cross-domain hoarding begin to emerge. Because our study is cross-sectional, we cannot determine causality or whether short-lived benefits transition into longer-term harms. Future research employing longitudinal latent transition methods (Vaziri et al., 2020) and experimental designs should (a) distinguish instrumental and social motives from compulsive motives, (b) track trajectories of performance, well-being, and addiction over time, (c) examine content and organization of hoarded materials, and (d) test moderators (e.g., self-regulation, digital literacy) predicting whether digital hoarding yields net benefit or harm. Clinically, this suggests future intervention research should focus on promoting adaptive curation and monitoring emerging addictive patterns rather than blanket reduction of saved material.

Conclusion

Research on hoarding behavior has focused on two types of behavior: digital- and physical-hoarding behaviors. To date, researchers have largely considered each behavior in isolation. Additionally, the variable-centered approach overlooks the possibility that subpopulations may differ in terms of their digital- and physical-hoarding behaviors; our study bridges this gap by using a person-centered approach to investigate the latent profiles associated with digital- and physical-hoarding behaviors. We conducted an LPA, and the results revealed 4 types of hoarding behavior. Our results suggested that digital hoarding can occur independently but also that physical hoarding can co-occur with digital hoarding. Moreover, these four profiles vary systematically depending on personal factors, and the profiles are associated with differences in important outcomes, including well-being, performance, and sleep.

Footnotes

ORCID iD

Haixia Wang

Ethical Considerations

The study involving human participants was reviewed and approved by the Ethics Committee of the first author’s university.

Consent to Participate

The participants signed informed written consent forms.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Program of National Natural Science Foundation of China (grant number 72174075), and Fundamental Research Funds for the Central Universities (23JNQMX56).

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The data referenced in this research are not publicly available at present as a result of ongoing research; however, they are available from the corresponding author upon reasonable request.*

References

Andreassen

C. S.

Torsheim

Brunborg

G. S.

Pallesen

(2012). Development of a Facebook Addiction Scale. Psychological Reports, 110, 501–517. https://doi.org/10.2466/02.09.18.PR0.110.2.501-517

Asparouhov

Muthén

(2014). Auxiliary variables in mixture modeling: Three-step approaches using Mplus. Structural Equation Modeling A Multidisciplinary Journal, 21, 329–341. https://doi.org/10.1080/10705511.2014.915181

Baer

M. D.

Dhensa-Kahlon

R. K.

Colquitt

J. A.

Rodell

J. B.

Outlaw

Long

D. M.

(2015). Uneasy lies the head that bears the trust: The effects of feeling trusted on emotional exhaustion. Academy of Management Journal, 58, 1637–1657. https://doi.org/10.5465/amj.2014.0246

Brand

Wegmann

Stark

Müller

Wölfling

Robbins

T. W.

Potenza

M. N.

(2019). The interaction of person-affect-cognition-execution (I-PACE) model for addictive behaviors: Update, generalization to addictive behaviors beyond internet-use disorders, and specification of the process character of addictive behaviors. Neuroscience and Biobehavioral Reviews, 104, 1–10.

Brislin

R. W.

(1980). Translation and content analysis of oral and written material. In Triandis

H. C.

Berry

J. W.

(Eds.), Handbook of cross-cultural psychology (pp. 389–444). Allyn & Bacon.

Büscher

T. P.

Dyson

Cowdell

(2014). The effects of hoarding disorder on families: An integrative review. Journal of Psychiatric and Mental Health Nursing, 21, 491–498. https://doi.org/10.1111/jpm.12098

Cath

D. C.

Nizar

Boomsma

Mathews

C. A.

(2017). Age-specific prevalence of hoarding and obsessive compulsive disorder: A population-based study. American Journal of Geriatric Psychiatry, 25, 245–255. https://doi.org/10.1016/j.jagp.2016.11.006

Charilaou

Vijaykumar

(2023). Influences of news and social media on food insecurity and hoarding behavior during the COVID-19 pandemic. Disaster Medicine and Public Health Preparedness, 17, PII S1935789321003153. https://doi.org/10.1017/dmp.2021.315

Gabriel

A. S.

Daniels

M. A.

Diefendorff

J. M.

Greguras

G. J.

(2015). Emotional labor actors: A latent profile analysis of emotional labor strategies. Journal of Applied Psychology, 100, 863–879. https://doi.org/10.1037/a0037408

10.

Grisham

J. R.

Barlow

D. H.

(2005). Compulsive hoarding: Current research and theory. Journal of Psychopathology and Behavioral Assessment, 27, 45–52. https://doi.org/10.1007/s10862-005-3265-z

11.

Grisham

J. R.

Williams

A. D.

Kadib

(2010). Hoarding as a behavioral addiction. In Johnson

(Ed.) Addiction medicine: Science and practice (pp. 687–701). Springer New York.

12.

Howard

M. C.

Hoffman

M. E.

(2018). Variable-centered, person-centered, and person-specific approaches: Where theory meets the method. Organizational Research Methods, 21, 846–876. https://doi.org/10.1177/1094428117744021

13.

Kim

Sung

Y. S.

McClure

S. M.

(2012). The neural basis of cultural differences in delay discounting. Philosophical Transactions of the Royal Society B: Biological Sciences, 367(1589), 650–656.

14.

Kim

(2013). Personal digital archives: Preservation of documents, preservation of self. http://hdl.handle.net/2152/21134

15.

Kreiner

G. E.

(2006). Consequences of work-home segmentation or integration: a person-environment fit perspective. Journal of Organizational Behavior, 27, 485–507. https://doi.org/10.1002/job.386

16.

Lanza

S. T.

Tan

Bray

B. C.

(2013). Latent class analysis with distal outcomes: A flexible model-based approach. Structural Equation Modeling, 20, 1–26. https://doi.org/10.1080/10705511.2013.742377

17.

Lin

Shao

Guo

Zhan

(2021). The psychological implications of COVID-19 on employee job insecurity and its consequences: The mitigating role of organization Adaptive Practices. Journal of Applied Psychology, 106, 317–329. https://doi.org/10.1037/apl0000896

18.

Mendell

N. R.

Rubin

D. B.

(2001). Testing the number of components in a normal mixture. Biometrika, 88, 767–778. https://doi.org/10.1093/biomet/88.3.767

19.

MacKenzie

S. B.

Podsakoff

P. M.

Fetter

(1991). Organizational citizenship behavior and objective productivity as determinants of managerial evaluations of salespersons’ performance. Organizational Behavior and Human Decision Processes, 50, 123–150. https://doi.org/10.1016/0749-5978(91)90037-T

20.

Maranci

J. B.

Aussel

Vidailhet

Arnulf

(2021). Grumpy face during adult sleep: A clue to negative emotion during sleep? Journal of Sleep Research, 30(6), e13369.

21.

Mesquita

(2001). Emotions in collectivist and individualist contexts. Journal of Personality and Social Psychology, 80(1), 68–74.

22.

Morgeson

F. P.

(2005). The external leadership of self-managing teams: Intervening in the context of novel and disruptive events. Journal of Applied Psychology, 90, 497–508. https://doi.org/10.1037/0021-9010.90.3.497

23.

Neave

Briggs

McKellar

Sillence

(2019). Digital hoarding behaviours: Measurement and evaluation. Computers in Human Behavior, 96, 72–77. https://doi.org/10.1016/j.chb.2019.01.037

24.

Neave

McKellar

Sillence

Briggs

(2020). Digital hoarding behaviours: Implications for cybersecurity. In Benson

Mcalaney

(Eds.), Cyber influence and cognitive threats (pp. 77–95). Elsevier. https://doi.org/10.1016/B978-0-12-819204-7.00005-1

25.

Nur-A Yazdani

D. M.

Abir

Qing

Ahmad

Al Mamun

Zainol

N. R.

Kakon

Agho

K. E.

Wang

(2022). Social media addiction and emotions during the disaster recovery period-The moderating role of post-COVID timing. PLoS One, 17, e0274898. https://doi.org/10.1371/journal.pone.0274898

26.

Nutley

S. K.

Bertolace

Vieira

L. S.

Nguyen

Ordway

Simpson

Zakrzewski

Camacho

M. R.

Eichenbaum

Nosheny

Weiner

Mackin

R. S.

Mathews

C. A.

(2020). Internet-based hoarding assessment: The reliability and predictive validity of the internet-based hoarding rating scale, self-report. Psychiatry Research, 294, 113505. https://doi.org/10.1016/j.psychres.2020.113505

27.

Peugh

Fan

(2013). Modeling unobserved heterogeneity using latent profile analysis: A Monte Carlo simulation. Structural Equation Modeling A Multidisciplinary Journal, 20(4), 616–639.

28.

Sedera

Lokuge

(2018). Is digital hoarding a mental disorder? Development of a construct for digital hoarding for future IS research [Conference session]. Proceedings of the 39th International Conference on Information Systems (ICIS 2018). University of Southern Queensland). https://aisel.aisnet.org/icis2018/behavior/Presentations/27/

29.

Shauly

Stone

Gould

(2020). The public’s perception of the severity and global impact at the start of the SARS-CoV-2 pandemic: A crowdsourcing-based cross-sectional analysis. Journal of Medical Internet Research, 22, e19768. https://doi.org/10.2196/19768

30.

Sillence

Dawson

J. A.

Brown

R. D.

McKellar

Neave

(2026). Digital hoarding and personal use digital data. Human–Computer Interaction, 41, 6–25.

31.

Sun

Lee

S. K.

(2022). Flooded with too many messages? Predictors and consequences of instant messaging fatigue. Information Technology and People, 35(7), 2026–2042.

32.

Sweeten

Sillence

Neave

(2018). Digital hoarding behaviours: Underlying motivations and potential negative consequences. Computers in Human Behavior, 85, 54–60. https://doi.org/10.1016/j.chb.2018.03.031

33.

Thorpe

Bolster

Neave

(2019). Exploring aspects of the cognitive behavioural model of physical hoarding in relation to digital hoarding behaviours. Digital Health, 5, 2055207619882172. https://doi.org/10.1177/2055207619882172

34.

Tolin

D. F.

Frost

R. O.

Steketee

(2010). A brief interview for assessing compulsive hoarding: The Hoarding Rating Scale-Interview. Psychiatry Research, 178, 147–152. https://doi.org/10.1016/j.psychres.2009.05.001

35.

Tolin

D. F.

Gilliam

C. M.

Davis

Springer

Levy

H. C.

Frost

R. O.

Steketee

Stevens

M. C.

(2018). Psychometric properties of the Hoarding Rating Scale-Interview. Journal of Obsessive-Compulsive and Related Disorders, 16, 76–80. https://doi.org/10.1016/j.jocrd.2018.01.003

36.

Valkenburg

P. M.

(2022). Social media use and well-being: What we know and what we need to know. Current Opinion in Psychology, 45, 101294.

37.

van Bennekom

M. J.

Blom

R. M.

Vulink

Denys

(2015). A case of digital hoarding. BMJ Case Reports. Advance online publication. https://doi.org/10.1136/bcr-2015-210814

38.

Vaziri

Casper

W. J.

Wayne

J. H.

Matthews

R. A.

(2020). Changes to the work–family interface during the COVID-19 pandemic: Examining predictors and implications using latent transition analysis. Journal of Applied Psychology, 105, 1073–1087. https://doi.org/10.1037/apl0000819

39.

Wang

Miao

Jia

Lai

(2023). The dark side of upward social comparison for social media users: An investigation of fear of missing out and digital hoarding behavior. Social Media + Society, 9, 20563051221150420. https://doi.org/10.1177/20563051221150420

40.

Watkins

M. B.

Ren

Umphress

E. E.

Boswell

W. R.

Triana

M. D. C.

Zardkoohi

(2015). Compassion organizing: Employees’ satisfaction with corporate philanthropic disaster response and reduced job strain. Journal of Occupational and Organizational Psychology, 88, 436–458. https://doi.org/10.1111/joop.12088

41.

Whelan

Najmul Islam

A. K. M.

Brooks

(2020). Is boredom proneness related to social media overload and fatigue? A stress–strain–outcome approach. Internet Research, 30, 869–887. https://doi.org/10.1108/INTR-03-2019-0112

42.

Woody

S. R.

Kellman-McFarlane

Welsted

(2014). Review of cognitive performance in hoarding disorder. Clinical Psychology Review, 34(4), 324–336.

43.

Huang

(2021). Development and validation of Digital Hoarding Behavior Scale. Psychology:Techniques and Application, 9, 116–125. https://doi.org/10.16842/j.cnki.issn2095-5588.2021.02.007

44.

Zhao

Cai

Gao

Liu

Wang

Zhang

Chen

Shi

(2022). Association between fear of COVID-19 and hoarding behavior during the outbreak of the COVID-19 pandemic: The mediating role of mental health status. Frontiers in Psychology, 13, 996486. https://doi.org/10.3389/fpsyg.2022.996486

45.

Zhong

Paluch

R. M.

Shum

Zatzick

C. D.

Robinson

S. L.

(2021). Hot, cold, or both? A person-centered perspective on death awareness during the COVID-19 pandemic. Journal of Applied Psychology, 106, 839–855. https://doi.org/10.1037/apl0000931