Binge watching and serial viewing: Comparing new media viewing habits in 2015 and 2020.

INTRODUCTION: The current study explores binge watching as a functional entertainment choice, and examines similarities and differences between it, appointment viewing, and serial viewing in terms of prevalence and technologies used over time. METHODS/MEASUREMENTS: Two surveys were conducted in fall 2015 (N = 373, 62% female and the mean age = 22.01 [SD = 5.92]) and in fall 2020 (N = 732, 69% female, mean age = 21.13 [SD = 4.98]. Surveys explored the frequency and duration of engaging in binge watching, serial viewing, and appointment viewing, as well as the role of habit in binge watching and the technologies employed to view content. FINDINGS/
RESULTS: Frequency of binge watching and serial viewing increased from 2015 to 2020, and levels remain elevated from estimates just prior to COVID-19 stay-at-home orders, which saw all types of viewing increase. Participants report most frequently engaging in serial viewing, followed binge watching, followed by appointment viewing. Appointment viewing frequency and duration has decreased. A binge-watching habit explains considerable variance in the frequency of binge watching.
CONCLUSIONS: Binge watching has become a more common way to watch TV than traditional appointment viewing, which has decreased in both the frequency and time spent between 2015 and 2020. Serial viewing - a self-paced consumption of serialized narrative content over days, weeks, or months - is the most common way of watching television content. Serial viewing and binge watching are more closely associated with viewing content on television, demonstrating a shift from the personal screen to the living room.

Introduction

While binge watching has quickly slipped into our lexicon and television viewing norms, empirical research on binge watching and other, evolving viewing behaviors is still quite new. A systematic review of binge-watching studies identifies the earliest empirical, quantitative article published in 2015, and a significant growth of publications beginning in 2017 (Flayelle et al., 2020). Much of this research has focused on defining binge watching, and assessing the degree to which it is either harmful or entertaining (Flayelle et al., 2017, Flayelle et al., 2019; Rubenking, Bracken, Sandoval, & Rister, 2018; Tukachinsky & Eyal, 2018). This study explores binge watching as a functional form of entertainment, one rooted in humans’ affinity for narratives. Binge watching is examined alongside traditional appointment viewing, and a third type of viewing that exists on a continuum between appointment and binge watching: Serial viewing. To explore differences in these qualitatively distinct viewing types, we employ cross-sectional surveys of U.S. university students in 2015 and 2020 examining the frequency and duration of engaging in these viewing types, the role of habit in binge watching, and changes in technologies used to view content.

The purpose of this study is twofold. First, by analyzing survey data collected in 2015 and 2020, this is the first exploration into whether viewing patterns, and binge watching specifically, have changed or have remained relatively the same since academics first started empirically studying them. Specifically, we examine the degree to which binge watching frequency can be predicted by an existing binge watching habit. Much has changed in five years, including a dramatic rise in the number of streaming options and an increase in cord-cutting from cable subscription. Additionally, the COVID-19 pandemic has had marked influence on television viewing habits. Netflix saw an unexpected influx of revenue, new subscribers, and time spent viewing in spring of 2020 (Alexander, 2020). One survey of participants from four Southeast Asian countries found that 73.7% reported an increase in time spent binge watching during the pandemic (Dixit et al., 2020). A survey of U.S. adults found that 52% reported spending more time streaming in summer 2020 than they would in a normal summer due to COVID-19 restrictions (Sadlier, 2020). Further, Pew Research reported that the vast majority of adults (9 -in-10) are watching media content to help them cope, with seventy-three (73) percent watching daily (Pew Research Center, 2020). The differences in the frequency of binge watching, and related viewing types, are thus important to isolate both between 2015 and 2020, and over the course of 2020.

Second, we compare three types of viewing including binge watching, serial viewing, and appointment viewing. Clarification of the different phenomenology associated with each mode of viewing is provided below.

Types of viewing

Binge watching

The lack of consistency in “binge watching” definitions and operationalizations makes the development of knowledge on the topic difficult. Flayelle et al.’s systematic review reports three common components of binge watching.: “(1) a quantity-based index; (2) the characterization of the content; and (3) a time pattern.”

We define binge watching as:

“Long periods of focused, deliberate viewing of sequential television content that is generally narrative, suspenseful, and dramatic in nature. Binge watching may be a planned, purposeful activity, or unintentional.” (Rubenking & Bracken, 2020).

Purposefully, this definition leaves out the number of episodes, as a basic plurality is needed. We instead focus on content that heightens anticipation or suspense. The deliberate viewing criterion emphasizes that this is a focused, attentive type of viewing. The time pattern is limited to one sitting, which we view as a critical distinction from serial viewing.

Serial viewing

The term serial comes from “serialization” of narrative fiction stories or the presentation of a story in pieces. This format of storytelling has strongly influenced how media content is presented to today’s audiences. O'Sullivan (2019) argues all forms of media storytelling today are shaped by the existence of a gap between episodes or serial installments. Television narratives employ particular mechanisms to bridge these gaps. Notably, they rely on a “heavy emphasis on character development and continuous storylines that flow between episodes of a series” (p. 23).

Serial viewing is defined as “Watching a series, a season, or several seasons of a TV show at one’s own pace over the course of several days, weeks, or months.” This encompasses much of the viewing that is done on streaming platforms, but does not encompass longer, singular binges. We think streaming service prevalence has assisted the ushering in of an important middle ground between binge watching and traditional appointment viewing of content. Indeed, some have noted the changes in serialized storytelling that the distributions strategies of companies like Netflix can encourage. McCormick (2016) identifies several ways in which House of Cards, an early hit for Netflix, made conscious content decisions based on viewers watching the shows in quicker succession, or without more traditional breaks between episodes. This included changing the placement of cliffhangers -- a major character’s death occurs in the season 2 premiere -- not a season finale, and a lack of recaps of previous episodes at the beginning of new episodes, and less repetition between episodes. Changes in distribution and streaming television content itself lends itself to a quicker consumption of content as compared to traditional television viewing, and not all of this sequential viewing is done in large chunks of time -- largely because of the zero-sum game of leisure time: There are so only many hours in a day.

Appointment viewing is the traditional viewing of content live or when it is broadcast, typically in a once-weekly format

A unique contribution of the current research is the ability to examine binge watching, as well as serial viewing and appointment viewing, over a 5-year period with a cross-sectional design. Therefore, the first research question is posed:

RQ1: What changes between the 2015 and 2020 samples will emerge for appointment viewing, serial viewing, and binge watching frequency and duration?

Media habits

Habit has long been overlooked in media use behaviors, though it explains significant portions of media behaviors (Adams, 2000). La Rose (2010) states that habit acquisition and strength is dependent upon greater repetitions, or opportunities to perform the behavior, and the ease with which one learns the association between the behavior and outcome. Researchers have previously found habit and automaticity to play a significant role in explaining binge watching (Rubenking and Bracken, 2018, Walton-Pattison et al., 2018). Media use habits may begin as goal-directed and beneficial, but may devolve into situations where self-control becomes ineffective, and behavior ceases to be goal-directed. This discrepancy between the findings may be due to conflating two distinct experiences of binge watching: Some have called for differentiating between intentional, highly involving binge watching experiences, and problematic binge watching (Flayelle et al., 2020, Riddle et al., 2018). Regardless of the functionality of a binge watching habit, we propose that binge watching habits are well ingrained among television viewers. Thus,

H1: Habit will strongly predict the frequency of binge watching in 2015 and 2020.

Media technology usage

Smart TVs and laptops are the most commonly used media technologies used to watch streaming television content. The majority of adults stream content using a smart television (30.8%) followed by mobile devices (16.3%), and then Amazon Fire (14.3%) and Roku devices (12.8%) (Stoll, 2020). Castro, Rigby, Cabral, and Nisi (2021) found young adults binge watched more often on their laptops as compared to television. Nielsen (2017) reports that the time people spend watching content on their laptops and mobile phones continue to increase. Americans are spending 11 h, 54 min each day consuming media, including television viewing, smartphones, radio, and gaming with simultaneous screen usage or second screening independently (Bauder, 2020). Determining context-specific displacement effects between devices and viewing types is of both theoretical and practical interest, and the ability to examine these changes over time is a novel opportunity. Therefore,

RQ2: What differences in frequency of viewing on different media technologies and frequency of engaging in each viewing type will emerge between the 2015 and 2020 samples?

Methods & measures

Sample and procedures

Cross-sectional online survey data was collected in two rounds: First from a student sample (N = 373) from two U. S. universities in fall, 2015, and second, from two U.S. universities in fall, 2020 (N = 732). The data from 2015 was combined with a sample of U.S. adults and analyzed together in Authors (2015). The demographic characteristics of the two samples were similar. In 2015, the sample was 62% female and M = 22.01 (SD = 5.92) years old. In 2020, the sample was 69% female and M = 21.13 (SD = 4.98) years old. All data were collected with approval from each institution’s Institutional Review Board. The survey took participants approximately 15 minutes to complete, and they earned extra course credit for their participation.

Measures

Types of viewing: 2015 and 2020

Participants were asked about both the frequency they engaged in three types of viewing, and how long, on average, those viewing sessions lasted. They were given the following definitions of Appointment Viewing, Serial Viewing, and Binge Watching:

Appointment Viewing: “Regularly viewing a TV show each week as it airs, live.”

Serial Viewing: “Watching a series, a season, or several seasons of a TV show at your own pace over the course of several days, weeks, or months.”

Binge Watching: “Watching 3–4 or more 30 min shows OR 3 episodes or more of hour-long TV shows in one sitting.”

Frequency of each type of viewing was assessed by a 9-point scale from “never” (1) to “a large part of every day” (9). The order of questions about each type of viewing was randomized for all participants. Participants were also asked, “On an average day that you (appointment view/ serial view/ binge watch) television content, how many hours do you watch for?” as a measure of the duration of viewing.

Binge watching habit

The four-item automaticity subscale of the Self-Report Habit Index, SRHI (Verplanken & Orbell, 2003) as previously used by Gardner et al. (2012) was used to index the extent to which binge watching was an existing habit. Grater values indicate a greater habit on this 7-point scale. The scale achieved excellent reliability in 2015 (α = 0.940) and in 2020 (α = 0.947).

Media technologies used to view television

The same 9-point scale (“never” (1) to “a large part of every day” (9)) measured the frequency of engaging in each type of viewing was employed to measure how often participants viewed television (in any type of viewing) on each of the following platforms: TV, laptop/ computer, and mobile device.

Frequency of types of viewing in 2020 -- the impact of COVID-19

Both anecdotal accounts and preliminary data suggest that the COVID-19 pandemic and accompanying stay-at-home orders served to increase the amount of time individuals spent viewing television, and specifically binge watching television (Dixit et al., 2020, Sadlier, 2020). We asked participants how often they engaged in each type of viewing at different time points in 2020. Participants were asked how frequently they engaged in each viewing behavior “just prior to covid-19,” “at the height of COVID-19 stay-at-home orders,” and “currently.” In the U.S., COVID-19 stay-at-home orders began in March 2020, and were the most restrictive and widespread measures put into place across the country during 2020. Thus, “at the height of COVID-19 stay-at-home orders” refers to approximately March and April of 2020, and “just prior to COVID-19 would refer to the month or so prior. The measures of viewing “currently” indicates the time period in which we collected the data: September 2020. It is this measure of frequency and duration of viewing (the latter of which was asked only of participants at the “currently” time point) that is used in hypothesis testing.

Results

COVID-19 and changes in TV viewing

Pairwise t-tests compared the frequency of engaging in each viewing type across the three time periods described above. Full results are presented in Table 1. The serial viewing and binge watching data show the same pattern over time: The greatest frequency of engaging in these viewing types was during the height of stay-at-home orders, followed by when data was collected in September 2020, followed by just prior to COVID-19. Participants indicated that they spent more time appointment viewing at the height of COVID-19 stay-at-home orders as compared to both just prior to COVID-19 and currently (the latter two of which did not significantly differ). These findings should not be strictly interpreted as a “new normal” of post-pandemic viewing frequency. September 2020 still found the U.S. (and the world) dealing with the pandemic and changes in day-to-day routine that could affect television viewing behaviors. Rather, this analysis shows a retrospective account of participants’ viewing habits changing throughout 2020.

Table 1

2020 comparisons of types of viewing frequency over time.

	A. Pre-COVID		B. Height of Stay-at-Home		C. Currently (September 2020)
Type of Viewing	M	SD	M	SD	M	SD	95% CI for Mean Difference	p	t	Effect size (d)
A. Appt. Viewing Frequency	3.27	2.11	3.72	2.53	3.20	2.30	−0.563, −0.336−0.045, 0.1740.399, 0.629	<0.0010.249<0.001	7.773*6.6668.772*	B > A (d = 0.29)A, C non-sig.B > C (d = 0.32)B > A, C
B. Serial Viewing Frequency	4.81	2.33	5.91	2.52	5.06	2.34	−1.223, −0.982−0.351, −0.1410.746, 0.967	<0.001<0.001<0.001	17.982*4.614*15.251*	B > A (d = 0.66)C > A (d = 0.17)B > C (d = 0.56)B > C > A
C. Binge Watching Frequency	4.22	2.17	5.94	2.41	4.50	2.30	−1.853, −1.592−0.403, −0.1710.066, 1/306	<0.001<0.001<0.001	25.881*4.864*21.796*	B > A (d = 0.96)C > A (d = 0.18)B > C (d = 0.81)B > C > A

Hypothesis and research questions

Research question 1 asked what differences will emerge for appointment viewing, serial viewing, and binge watching frequency and duration between the 2015 and 2020 samples. In order to answer this question, the 2015 and 2020 datasets were merged so that the year of data collection could be entered as the predictor variable in six one-way ANOVAs to determine differences in binge watching, serial viewing, appointment viewing frequency and duration. No post hoc tests were needed as each ANOVA tested the differences between just two means.

In terms of the frequency of engaging in each viewing type, data analysis reveals that the largest change, and the only decrease in frequency from 2015 (M = 4.45, SD = 2.26) to 2020 (M = 3.20, SD = 2.30) was found for appointment viewing F(1, 1101) = 86.996, p < .001. Both serial viewing and binge watching saw modest jumps in frequency from 2015 to 2020. Serial viewing saw the smaller increase from 2015 (M = 4.50, SD = 2.20) to 2020 (M = 5.06, SD = 2.33, F(1, 1079) = 13.970, p < .001. While binge watching saw only a slightly larger increase from 2015 (M = 3.75, SD = 1.95) to 2020 (M = 4.50, SD = 2.30, F(1, 1075) = 27.600, p < .001.

The average duration of each session of viewing television among all three types experienced small decreases from 2015 to 2020, with the smallest change found in serial viewing from 2015 (M = 2.47, SD = 2.16) to 2020 (M = 2.17, SD = 2.04, F(1, 1062) = 4.827, p = .028. Changes in average duration of viewing were similar across appointment viewing in 2015 (M = 2.04, SD = 2.51) and 2020 (M = 1.26, SD = 1.95, F(1, 1094) = 31.646, p < .001) and in binge watching in 2015 (M = 4.12, SD 2.53) and 2020 (M = 3.40, SD = 2.22, F(1, 1055) = 21.611, p < .001. In response to RQ1, analysis of data reveals increases in the frequency of serial viewing and binge watching, and a decrease in the frequency of appointment viewing from 2015 to 2020. All three types of viewing saw significant decreases in the duration of each type of viewing session, although serial viewing remained the most similar over time.

Hypothesis one predicted that habit would explain a significant amount of variance in binge watching frequency in 2020, as it had in 2015. Multiple regressions on binge watching frequency were run in each, independent, dataset. Age and sex were controlled for in the first block, and habit was entered in the second block. Standardized variables were employed. The regression utilizing the 2015 dataset was statistically significant, R = 0.369, F(2, 335) = 54.802, Adj. R = 0.323, p < .001. The first block of the regression was not significant, but the second block was, where Habit’s β = 0.567, p < .001. There was an independence of residuals, as assessed by a Durbin-Watson statistics of 2.050. There were also no signs of multicollinearity issues, with VIF values under 1.03. Results were similarly significant with the 2020 data: R = 0.329, F(2, 725) = 141.370, Adj. R = 0.366, p < .001. Independence among variables was found (Durbin-Watson = 1.912) and there were no multicollinearity issues (VIFs < 1.02). Habit also demonstrated a significant Beta weight, β = 0.602, and neither age nor sex were significant in the final block. Hypothesis 1 is supported: Habit was a large predictor of binge watching frequency in 2020, and grew marginally from the 2015 numbers.

The second research question asked about the relationship between the frequency of viewing television on different devices and the frequency of engaging in each of the viewing types. Correlations between the frequency of technology use and frequency of viewing type are presented in Table 2, Table 3. The data reveal a relatively stable story for binge watching and appointment viewing. Binge watching is associated with greater use of all three technologies to view television in both 2015 and 2020. Notably, the largest change in correlations between frequency of technology used and binge watching was the growth from r2 = 0.142, p < .001 between binge watching and TV use in 2015 to r2 = 0.328, p < .001 in 2020. In 2015 and 2020, appointment viewing is more greatly associated with viewing TV, on TV, which seems quite logical. It also shares a small correlation with mobile viewing in 2015. Serial viewing presents a more dynamic relationship with technologies used. While consistently associated with greater use of a laptop to view content, it shares a small positive relationship with mobile viewing in 2015, but does not in 2020. Instead, we see a large, moderate-sized correlation emerge between serial viewing frequency and TV viewing frequency in 2020 (r2 = 0.356).

Table 2

2015 frequency of viewing TV on different media technologies across different viewing types.

	1	2	3	4	5	6
1. TV on TV	1
2. TV on Computer	−0.039	1
3. TV on Mobile	0.153*	0.370**	1
4. Frequency Appointment Viewing	0.400**	−0.023	0.136*	1
5. Frequency Serial Viewing	0.087	0.241**	0.155*	0.129*	1
6. Frequency Binge Watching	0.143**	0.214**	0.174**	0.195**	0.465**	1

Table 3

2020 frequency of viewing TV on different media technologies across different viewing types.

	1	2	3	4	5	6
1. TV on TV	1
2. TV on Computer	−0.114*	1
3. TV on Mobile	0.013	0.296**	1
4. Frequency Appointment Viewing	0.264**	−0.020	0.057	1
5. Frequency Serial Viewing	0.356**	0.110*	0.066	0.211**	1
6. Frequency Binge Watching	0.328**	0.215**	0.215**	0.125*	0.553**	1

Discussion

Turner (2021) echos others in media studies when he states that binge watching has “outlived its usefulness for television studies” (p. 229), citing it’s numerous conceptualizations and operationalizations. We believe this empirical study works to narrow the definition of binge watching, and provides a solid rationale for identifying a new viewing style made possible by the wide availability of streaming services that is a not-quite binge watching, but not quite traditional TV watching: Serial viewing. We see evidence of this across the results reported here. Serial viewing increased in frequency from 2015 to 2020, making it more frequently engaged in than both appointment viewing and binge watching in 2020. The data on the duration of serial viewing puts it squarely in between longer binge sessions, and shorter appointment viewing sessions. And it’s smallest relative increases in frequency and decreases in duration, as compared to the other two types of viewing, demonstrating a consistent, stable type of viewing television. Binge watching remains a more frequently engaged in mode of viewing than appointment viewing, which saw the largest decreases in frequency and duration among the three viewing types.

Serial viewing has its own, distinct patterns of frequency across the 2015 and 2020 datasets, as well across media technologies as compared to binge watching. The association between serial viewing (and binge watching) and viewing on a television has grown from 2015 to 2020, which speaks to the normalization and creation of new media habits of these types of viewing. Additionally, the increase of binge watching on television sets suggests that more members of households are engaging in binge watching contrary to the assertion by Castro et al (2021), possibly as part of a group. We also saw the penetration rate of Internet connected or Smart TVs increase during this time (Vorhaus, 2020).

The stability of habit’s considerable influence on binge watching frequency, along with the consistent binge watching frequency and duration data, suggest that binge watching has become a new media habit that remains an intentional, highly involving viewing experience rather than a problematic binge watching experience for most audience members (Flayelle et al., 2020, Riddle et al., 2018).

The normalization of binge watching and serial viewing suggest that future media studies should measure the viewing styles of audiences when investigating media effects. One limitation of the current study was the use of college students. The participants were asked to self-report on their amount of viewing and self-report data is not always the most reliable, especially when asked about behaviors that took place several months earlier, seen in the COVID-19 data.

Conclusion

Binge watching is becoming a more common way to watch television content than traditional appointment viewing, which has decreased in both the frequency and time spent between 2015 and 2020. The increased frequency in serial viewing and binge watching that occurred at the height of the spring 2020 stay-at-home orders has remained elevated from just prior to the pandemic. Habit is a strong predictor of binge watching frequency, and future research should explore habit as related to serial and appointment viewing. Serial viewing, along with binge watching, has now become the most common viewing styles for people with access to streaming services or over-the-top, OTT, content providers. Newer viewing styles are not confined to any one media technology format, indicating binge watching has transcended from the personal screen to the living room.

Credit authorship contribution statement

Bridget Rubenking: Conceptualization, Methodology, Formal analysis, Visualization, Writing - original draft, Writing - review & editing. Cheryl Campanella Bracken: Conceptualization, Visualization, Writing - original draft, Writing - review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.