Third-party online organizational reviews: Explaining review-and-rating patterns of the United States military and large corporate organizations

Virtual spaces are a new and influential means by which present and past organizational members share reviews of their organizational experiences and socialize potential newcomers; however, online reviews can be negative and jeopardize an organization’s image. This investigation employed social identity theory and uncertainty management theory as a means of explaining patterned user ratings of organizational reviews online. In a first study, we content analyzed socialization storytelling about Basic Training on americangrit.com. Statistical analysis revealed that viewers rated stories more highly when the story portrayed the military favorably. In a second study, a content analysis of organizational reviews posted to indeed.com replicated and extended this pattern: Website visitors rewarded positive reviews of U.S. Military branches with higher ratings, while reviews of large corporate organizations (i.e., Apple, Bank of America, Michelin) varied. Implications for theory and practice conclude the paper. Attracting and retaining skilled employees is key to organizational success (Pfeffer & Veiga, 1999). Organizations spend valuable resources to attract, train, and keep employees. However, organizational image management is no longer solely in the hands of careful crafters of recruitment messages but also with anyone who uses the internet. Virtual spaces, such as indeed.com and americangrit.com, supply recruitment forums where individuals can post organizational reviews, anonymously, about an organization. Thus, the advent of the digital age and digital natives (i.e., young adults who have never known the absence of personal computing) has meant socialization messages can now be communicated across time and space in ways not previously possible. To date, studies of organizational socialization have primarily focused onmessages received from parents, teachers, and other influencers prior to entering the workforce (e.g., Scarduzio, Real, Slone, & Henning, 2018). The present study adds to these investigations the observation that virtual spaces are a new venue through which potential, present, and past members swap socialization information. Within this new context, this paper reports an investigation of two virtual spaces: americangrit.com and indeed.com. A first content analytic study investigates one section of americangrit.com, titled, Awesome Sh*t My Drill Sergeant Said (ASMDSS). The section is devoted to the swapping ofile/William_Howe8. ofile/Ryan_Bisel. 17 February 2020; Accepted 17 cess article under the CC BY lice of socialization experiences at Basic Training or Boot Camp. An analysis of the military offers a chance to examine a strict, totalistic, value-laden organizational form that will likely have similar uncertainty and identity management needs among potential, past, and present members (explained in detail below). The present investigation is also distinct from previous organizational studies of virtual spaces for at least three reasons: First, unlike virtual spaces used exclusively for exchanging critiques of an organization (e.g., radioshacksucks.com; Gossett& Kilker, 2006), ASMDSS focuses thematically on the exchange of stories about a significant and shared organizational entry message (i.e., “Drill Sergeants Said”). Second, and similarly, unlike virtual spaces used exclusively for venting and backstage dissent, ASMDSS contains negative, neutral, and positive storytelling—a point described in more detail below. Third, the patterns of organizational reviews-and-user rating between the U.S. Military and traditional organizations are evaluated in a follow-up study of a thematically-neutral virtual space (i.e., indeed.com). In sum, these studies contribute to the literature the idea that producers and consumers of military reviews systematically reward positive organizational portrayals with favorable ratings, whereas that pattern of review-and-rating were unrelated or negative when it involved traditional, large corporations. Additionally, we contribute to the organizational socialization literature

Virtual spaces are a new and influential means by which present and past organizational members share reviews of their organizational experiences and socialize potential newcomers; however, online reviews can be negative and jeopardize an organization's image. This investigation employed social identity theory and uncertainty management theory as a means of explaining patterned user ratings of organizational reviews online. In a first study, we content analyzed socialization storytelling about Basic Training on americangrit.com. Statistical analysis revealed that viewers rated stories more highly when the story portrayed the military favorably. In a second study, a content analysis of organizational reviews posted to indeed.com replicated and extended this pattern: Website visitors rewarded positive reviews of U.S. Military branches with higher ratings, while reviews of large corporate organizations (i.e., Apple, Bank of America, Michelin) varied. Implications for theory and practice conclude the paper.
Attracting and retaining skilled employees is key to organizational success (Pfeffer & Veiga, 1999). Organizations spend valuable resources to attract, train, and keep employees. However, organizational image management is no longer solely in the hands of careful crafters of recruitment messages but also with anyone who uses the internet. Virtual spaces, such as indeed.com and americangrit.com, supply recruitment forums where individuals can post organizational reviews, anonymously, about an organization. Thus, the advent of the digital age and digital natives (i.e., young adults who have never known the absence of personal computing) has meant socialization messages can now be communicated across time and space in ways not previously possible. To date, studies of organizational socialization have primarily focused on messages received from parents, teachers, and other influencers prior to entering the workforce (e.g., Scarduzio, Real, Slone, & Henning, 2018). The present study adds to these investigations the observation that virtual spaces are a new venue through which potential, present, and past members swap socialization information.
Within this new context, this paper reports an investigation of two virtual spaces: americangrit.com and indeed.com. A first content analytic study investigates one section of americangrit.com, titled, Awesome Sh*t My Drill Sergeant Said (ASMDSS). The section is devoted to the swapping of socialization experiences at Basic Training or Boot Camp. An analysis of the military offers a chance to examine a strict, totalistic, value-laden organizational form that will likely have similar uncertainty and identity management needs among potential, past, and present members (explained in detail below). The present investigation is also distinct from previous organizational studies of virtual spaces for at least three reasons: First, unlike virtual spaces used exclusively for exchanging critiques of an organization (e.g., radioshacksucks.com; Gossett & Kilker, 2006), ASMDSS focuses thematically on the exchange of stories about a significant and shared organizational entry message (i.e., "Drill Sergeants Said"). Second, and similarly, unlike virtual spaces used exclusively for venting and backstage dissent, ASMDSS contains negative, neutral, and positive storytelling-a point described in more detail below. Third, the patterns of organizational reviews-and-user rating between the U.S. Military and traditional organizations are evaluated in a follow-up study of a thematically-neutral virtual space (i.e., indeed.com). In sum, these studies contribute to the literature the idea that producers and consumers of military reviews systematically reward positive organizational portrayals with favorable ratings, whereas that pattern of review-and-rating were unrelated or negative when it involved traditional, large corporations. Additionally, we contribute to the organizational socialization literature Organizational socialization is defined as "the methods and processes by which new and continuing members of organizations 'learn the ropes'" (Jablin, 1985, p. 261). The topic has been popular in organizational communication studies for decades (for a review, see Kramer, 2010). Organizational socialization is a communication process through which members learn organizational norms and routines (Jablin, 2001;Kramer & Dailey, 2019). Many scholars noted that potential members and newcomers are motivated by anxiety and uncertainty to seek out socialization messages, which can aid them in acquiring social knowledge needed to signal group inclusion in their words and actions (Kramer, 2010;Kramer & Dailey, 2019). Over the years, several models of organizational socialization have been proposed. Jablin (2001) synthesized these models into a four-stage process, involving anticipatory socialization, encounter, metamorphosis/role management, and exit. Anticipatory socialization refers to those messages that create expectations about the meaning of joining an organization. Encounter socialization refers to those early membership message exchanges through which members learn the particulars of their new organization and role within it. Metamorphosis socialization describes those messages that reinforce members' status as fully established within the organization. Finally, exit refers to the dissolution of membership, either through voluntary or involuntary means (Herrmann, 2017).
Organizational socialization is a popular concept in the organizational communication literature; however, the concept is not a theory per se. Instead, Kramer (2010) observed that scholars tend to employ one of four theories when investigating this workplace dynamic: social identity theory, uncertainty management theory, social exchange theory, and sensemaking theory. For the present study, social identity theory (SIT) and uncertainty management theory (UMT) are particularly explanatory. The rise of online venues for sharing socialization information among potential, present, and past members begs the question: "What patterns of messaging and audience response can we expect in this emerging venue?" In the following paragraphs, SIT and UMT are explained and used to anticipate the answer to this question.
Social identity theory. Tajfel and Turner's (1985) social identity theory explains that individuals are motivated to answer the existential question, "Who am I?" To do so, they rely, in part, on salient group memberships to make sense of their self-concept (Postmes & Branscombe, 2010;Stets & Burke, 2000). The idea is significant in that it proposes individuals think and experience their world in terms of the salient group memberships they claim. Once a group membership becomes integrated into the self-concept, individuals can vicariously partake in the accomplishments of the group and will work to see the group in favorable terms (Brown & Starkey, 2000;Ploeger & Bisel, 2013). In turn, individuals protect positive notions of the self. Thus, SIT implies that a goal of communicators is to process information selectively such that potential and current salient group memberships are bolstered (Keyton, 2014).
Uncertainty management theory. Kramer (1993Kramer ( , 2010 reconceptualized Berger and Calabrese's (1975) uncertainty reduction theory. Kramer's uncertainty management theory (UMT) explains that individuals are motivated to navigate anxiety that is created by uncertainty associated with their indeterminate future as an organizational member. Thus, individuals seek out information, especially when they believe information will reduce their anxiety and help them learn the norms of the workplace. At other times, individuals will avoid seeking information or attempt to maintain uncertainty in their workplace communication, especially when they believe information and clarity will exacerbate their anxiety about the future. Thus, UMT implies that a goal of communicators is to manage anxiety about the future through seeking or avoiding information, depending on the situation. Therefore, in terms of organizational socialization, SIT and UMT suggest that potential, present, and past organizational members may be differentially motivated in how they produce and consume socialization information. According to SIT, potential members may look for organizational membership information, which bolsters the image of those organizations they are likely or eligible to join, while also seeking out information that denigrates the image of those organizations that they are not likely or eligible to join, during a process referred to as "cybervetting" (Berkelaar, 2013, p. 36). Similarly, UMT suggests that potential members, who have not yet been formally offered membership in an organization, will be open to a mix of information, both positive and negative to reduce the anxiety associated with their decision to join or not. However, once potential members decide that joining a group is possible and would be a positive addition to their self-concept, then they will be motivated to seek out information that bolsters the image of the organization and avoid, or discount, information that denigrates it. Now consider the motivations of present members from a SIT perspective: Present members will tend to seek out information that bolsters the meaning of their current membership and, by extension, their own self-concept. Likewise, UMT suggests that present members are motivated to seek out positive and avoid negative information thereby minimizing anxieties related to their role, membership, organization, or profession. Thus, the patterns of communication motivation described by both SIT and UMT are likely similar for present members. Finally, consider the motivations of past members from a SIT perspective: Past members may be ambivalent in their need to seek out information that bolsters the image of organizations with which they left; on the other hand, past members may be motivated to seek out information that denigrates the image of organizations from which they exited after relational dissolution. Meanwhile, UMT again leads us to believe that past members will seek out information that helps them to make sense of their uncertainties related to exiting. For those members who exited on positive terms or who wish to retain an identity-connection, UMT proposes that these members will produce positive organizational portrayals and accounts of their former roles within the organization. Conversely, those members who do not wish to retain an identity-connection to their former organization may avoid information or produce and consume negative information to reduce any anxiety associated with this relational dissolution. These members may also be motivated to reduce anxiety by reconciling that their former organization is inferior and then to communicate this belief to others as a means of identity protection.
These varying motivations for both producing and consuming information may contribute to mixed online organizational reviews, in terms of presenting organizations' favorably. However, membership in the military may create a different dynamic: Technically, military organizations are known as totalistic organizations in that they enact strong control over members both inside and outside the normal context of organizational work (Howe & Hinderaker, 2018). Additionally, membership in the U.S. Military is somewhat unique in that professional identification and organizational identification align strongly in the sense that the profession of soldier, sailor, airman (sic), or marine cannot be performed in other organizations (within the U.S.). Thus, military members and veterans tend to be highly identified and susceptible to inculcation such that membership becomes significantly enmeshed with the self-concept (Howe & Shpeer, 2019). In this way, potential, present, and past members of the military might be especially motivated to bolster their organization's image when producing and consuming socialization information.
Totalistic organizations (TOs) are organizational systems that integrate virtually all aspects of member's lives within their organizational membership and role (Hinderaker, 2015(Hinderaker, , 2017. To date, most totalistic organizational scholarship investigates the particularly consuming nature of membership within religious organizations (Bromley & Melton, 2002). Only recently have scholars examined such organizations from a communication perspective (e.g., Garner, 2016;Garner & Peterson, 2017;Shpeer & Howe, 2020;Tracy, 2000). Hinderaker (2015) stated that in a totalistic organization "values, practices, rituals, and relationships associated with membership not only extend into the member's everyday life but play a primary role" (p. 93). Thus, socialization into TOs are often marked by extreme transformations of identity, thought, and action (Howe & Hinderaker, 2018).
Hinderaker (2015) argued high-reliability organizations (HROs) will often be totalistic in nature because of the need for collective mindfulness (Weick, Sutcliffe, & Obstfeld, 2005), group unity created by rigorous training (Shpeer & Howe, 2020), and team cohesion under life-threatening pressures (Jahn & Black, 2017). In fact, Howe and Hinderaker (2018) documented how members of the United States Military and members of the ultra-conservative Independent Fundamental Baptist Church described similar organizational entry socialization experiences. Thus, the military is almost certainly a TO and its socialization experiences-such as those experienced during initial training-are likely to have strong identity-and value-shaping implications. Notably each branch of the military lists a set of core values that recruits must learn during initial training and these values influence recruits' behavior after they become a member and even after they leave the organization (Howe & Shpeer, 2019;Smith & True, 2014) similar to what has been found in other TOs (e.g., Hinderaker, 2015).

Totalistic organizations and virtual spaces
Howe and Hinderaker (2018) found that the intensely-structured process of TO entry facilitates a rapid stripping of individual identity and replacement with military identity. Knight (1990) previously explored one mechanism through which this transformation occurs by examining the cadence songs used by military leaders to motivate individuals during physical training. He found that these songs both normalized and relieved tension associated with basic training. Knight summarized his findings poignantly in the following paragraph: By laughing at the unpleasant realities of war, we no doubt were hardening ourselves to our own squeamishness and fear. Such hardening was to make us efficient soldiers, willing to kill or die on command (and, as officers. willing to give such commands). All of the attitudes I have discussed are geared toward this goal. The military mindset assumes that a soldier should, under certain conditions, hold no reservations about killing women or children, should function not as an individual but as part of a unit, and should feel absolute loyalty to the service. (p. 166) Knight's examination of cadence calls has the implied result of being identity transformative. In the present study, we examine the information about this entry process that gets circulated online via online stories and reviews.
Recent research in communication theorized (Berkelaar, 2013;Kramer, 2010) and documented empirically (Dailey, Treem, & Ford, 2016;Kramer, Lee, & Guo, 2019;Piercy & Lee, 2018) that potential organizational members are increasingly relying on virtual spaces as a means of information seeking in combination with traditional methods. Flanagin and Waldeck (2004) propose that a variety of socialization, individual, and normative factors will influence what types of media technology potential members and newcomers use during the socialization process. The authors view these technological advancements and ease of finding information positively in the sense that readily-available information about organizations and jobs will likely lead to a better fit between employee and organization (Waldeck, Seibold, & Flanagin, 2004). Recently, Knobloch et al. (2017) found that military members turned to virtual spaces to reduce relational uncertainty and seek social support. Due to the affordances of virtual spaces, as a means of managing uncertainty, potential members may turn to virtual spaces to reduce their uncertainty about joining. In fact, scholars noted for years that individuals are motivated to seek information selectively to manage their uncertainties and anxieties about joining a workplace (e.g., Kramer, 1993); presumably, virtual spaces are just a new forum for members' selective consumption of socialization information (Kramer, Lee, & Guo, 2019;Piercy & Lee, 2018;Wilson et al., 2019).
Importantly, reviews and descriptions on virtual spaces are not necessarily positive socialization information. For example, Gossett and Kilker (2006) examined a counter-institutional website, radioshacksucks.com, where upset customers or employees would leave negative reviews and comments about the organization. The apparent purpose of radioshacksucks.com was to disparage Radio Shack-a purpose that eventually motivated the corporation to litigate against the website for liable. The corporation won their suit and the site was shut down by court order. Although some messages on radioshacksucks.com may have been pro-institutional, that was certainly not the purpose of the site, as indicated by its namesake. Thus, individuals may use virtual spaces to circulate and consume dissent messages about organizations (Sanderson, 2009) and engage in socialization practices (Berkelaar, 2013). Yet, the cases of americangrit.com and indeed.com illustrate how potential, present, and past members produce, disseminate, and consume counter-institutional as well as pro-institutional and neutral information. As virtual spaces become increasingly common and available, individuals may seek out and consume these messages thereby changing the nature of socialization in the 21st century.

Study 1
The virtual space, ASMDSS, is a forum for users to produce and disseminate review stories about organizational entry found on americangrit.com. Other users-who could be potential members of the U.S. Military-may consume these stories as a part of their anticipatory socialization and as a means of reducing anxiety and uncertainty associated with joining. See Table 1 for a comparison of organizational review site features. Notably, ASMDSS is a virtual space that represents a unique opportunity to investigate the patterns of user ratings regarding detailed reviews of a specific socialization experience.
Americangrit.com features several categories including short humorous videos, military-related news articles, clothing advertisements, as well as several repositories of user-generated review stories about their military experiences. One category of review stories involves the prompt, "Basic Training Stories," alongside a logo that features a skull wearing a U.S. Army Drill Sergeant hat and smoking a cigar with the letters "ASMDSS" underneath. The acronym is recognizable to military members as meaning "Awesome Sh*t My Drill Sergeant Said" (ASMDSS); importantly, the acronym is not defined on the site, suggesting that story contributors readily interpreted the image and acronym and demonstrated familiarity with military culture and jargon.
In the process of reading the site, we noticed that stories varied in terms of how favorably the military and Drill Sergeants (DS) were portrayed. Some stories bolstered the image of the military and DS, while others were harder to discern, ambivalent, or unflattering. That observation prompted us to theorize about the nature of stories on this virtual space and how viewers rated those stories. We reasoned that potential newcomer members would be motivated to favor stories that presented the military and DSs positively because, in doing so, their anxiety and uncertainty about joining would be reduced and soothed. Similarly, present members identity and uncertainty management needs would motivate them to reaffirm positive stories and disparage or avoid negative ones. Additionally, we reasoned that past members would be motivated to favor stories that presented the military and DSs positively because, in doing so, many visitors would vicariously affirm themselves and their salient group identity of being a veteran. In other words, this line of reasoning led us to theorize that when the identity and uncertainty needs of potential, present, and past members align strong patterns of online reviewer ratings will be apparent. Thus, it was hypothesized: H1. Review stories that portray the military positively are positively associated with virtual space viewer ratings, even after controlling for number of story views, length of story, and days elapsed.
The control variables of story views, length of story, and days elapsed were used to account for the relationships between these potential covariates and viewer ratings. The addition of these control variables is important for ruling out alterative explanations. For example, it is possible that the website's governing algorithm rewards clicked stories with a more prominent placement on the website and, in turn, generates more story views and higher ratings. Controlling for number of story views will eliminate the possibility that findings are the result of a website algorithm and not the hypothesized relationship. Likewise, it is also possible that long stories are more beautifully or dramatically written and may therefore receive higher ratings. Again, controlling for story length will rule out the possibility that findings are merely the result of this pattern. It is also possible that stories that were posted at a certain period in time or for a longer period of time received more favorable ratings than others because of concurrent events or familiarity that have nothing to do with the story content. Therefore, these variables were controlled for as potential covariates, but no hypotheses were made as to their direction.

Study 1: method
At the time of data collection approximately 2,000 review stories were viewable on the forum. Two hundred stories (n ¼ 10%) were randomly selected from this corpus, via a random number generator (Social Psychology Network, randomizer.org). Each review story had ratings from site visitors on a 1-5 "star" scale. Those ratings were captured along with story length, days elapsed, number of views, and number of raters. As an inclusion criterion, review stories had to have been rated by at least 10 raters to ensure scores were supported by multiple website visitors. Review stories ranged in length from 126 to 1,099 words (M ¼ 338.88; SD ¼ 169.73); days elapsed ranged from 85 to 418 days (M ¼ 228.89; SD ¼ 90.94). The number of views ranged from 1,600 to 25,500 (M ¼ 9,460; SD ¼ 3,020)-a descriptive statistic which indicated the relative popularity of the website.
In order to test H1, each review story was coded in terms of how it portrayed the military, ranging from 1 (strongly negative), 2 (weakly negative), 3 (neither or neutral), 4 (weakly positive), to 5 (strongly positive; see Table 2 for a codebook). We relied on one of the author's military training and experiences to interpret military acronyms, cultural references, and jargon. Review stories that were coded "strongly negative" were often represented by disloyalty to the military, drill sergeants/ instructors, or other military members. "Weakly negative" stories were represented by similar features but were not as severe or overt in terms of disloyalty. Stories that demonstrated "neither or neutral" were rare and tended to be poorly-written review stories. Stories that exhibited "weakly positive" portrayals were often represented by moderate loyalty. Stories that demonstrated a "strongly positive" portrayal were represented by strong loyalty to the military, drill sergeants/instructors, or other military members.
Two raters independently rated all stories. To enhance rating validity, raters were not privy to viewer ratings. Intercoder reliability was sufficiently high (Krippendorff's alpha ¼ .81). Additionally, seven military veterans, who were blind to the study's objective, were asked to rank a selection of the ASMDSS review stories in terms of how favorably each portrayed the military. The untrained coders' rankings were consistent with the trained coders' ratings, suggesting the coding scheme was semantically valid (Krippendorff's alpha ¼ .88).

Study 1: results
Descriptive statistics and a correlation matrix among study variables are reported in Table 3. Story portrayal was significantly and positively related to visitor ratings (r ¼ .40, p < .001), however, other variables were also significantly related with story portrayal. Given that the story portrayal existed before the ratings and that the story portrayal covaried with the ratings, regression helped determine the most influential variable on visitor ratings. Regression analysis revealed the favorability of review story was significantly and positively predictive of website visitor ratings (β ¼ .41, p < .001), even after controlling for the influence of story views, length of story, and days elapsed (Table 4). In other words, website visitors tended to reward those stories that portrayed the military in a positive manner with high ratings. Thus, H1 was supported. Importantly, correlational analysis also revealed that the favorability of military portrayal was positively, but weakly, associated with story views (r ¼ .15, p < .05). Additionally, website user rating of story was positively, and strongly, associated with story views (r ¼ .49, p < .001). That relationship seems to indicate that favorably-valanced and highly-rated stories attracted the attention of viewers and, in turn, tended to receive more views.

Study 1: discussion
Study one provides support for H1 in that the favorability of military portrayal in review stories was significantly associated with positive user ratings. That finding is consistent with our theorizing that the identity and uncertainty needs of potential, present, and past members of the United States Military align and explains why positive organizational reviews tend to be rewarded with favorable ratings on the website. Additionally, stories that portrayed the military favorably received significantly more views than stories that portrayed the military negatively, although the association was small. Although not hypothesized, this finding is also consistent with social identity theory's claim that members seek to protect their self-concept by preferring information that bolsters the image of salient group membership. Furthermore, this finding is consistent with uncertainty management theory's claim that organizational members will seek information that reduces anxiety while avoiding information that exacerbates it. However, another possible alternative explanation exists: Perhaps readers were influenced by the word "Awesome" in the title of the website.

Study 2
Website visitors may have been primed by the title to expect to swap positive review stories and their expectations were violated by any reviews that were negative. On the other hand, the title of website also includes a presumably negative label (Sh*t). Nevertheless, a second study was conducted on another anonymous organizational review site, indeed.com, which has the benefit of being titled in a thematically-neutral manner (see Table 1). Another benefit of the website was its ability to offer the chance to explore whether the pattern was apparent of the three main branches of the U.S. Military (the Marines are technically a subbranch of the Navy). We hypothesized that the same relationship would be identified on indeed.com as on americangrit.com.
H2. Organizational reviews favorability of the (a) Air Force, (b) Army, and (c) Navy are positively associated with virtual space viewer ratings.
In addition to attempting to replicate the pattern of review-and-rating identified on ASMDSS, indeed.com also offered the opportunity to explore this relationship regarding reviews of traditional, large corporations. From a theoretical perspective, potential, present, and past members of traditional large corporations may, presumably, have differing identity and uncertainty needs as compared to potential, present, and past members of a totalistic organization, such as the U.S. Military. For example, unlike with the professions of soldier or sailor, a banker can perform their profession at many different financial institutions. More options for joining and exiting present varying identity and uncertainty challenges and attendant needs, which should complicate the favorable review-positive rating pattern identifiable among online reviews of the U.S. Military. We therefore asked:  .14*** .30*** ΔR 2 -.16*** Note: Model 1: RQ1: How are organizational reviews of (a) Apple, (b) Bank of America, and (c) Michelin associated with virtual space viewer ratings, if at all?

Study 2: method
The site indeed.com was chosen as it (a) provides reviews of multiple organizations, including all branches of the military as well as numerous traditional organizations, (b) reviewers provide their own rating of the organization on a 1-5 scale, and (c) viewers of these reviews can rate these reviews as helpful or not helpful (see Table 1).

Data collection
In order to test H2, we downloaded the first 100 reviews, which had a minimum of 10 user ratings, for the three main branches of the military (Army, Air Force, and Navy).
In order to answer RQ1, we sought to identify reviews of comparable traditional organizations (as opposed to a totalistic organization, such as the U.S. Military). To that end, traditional organizations had to meet the following criteria to be included in the analysis: First, traditional organization for comparison must have at least 100 reviews on indeed.com. Second, they needed to represent a range of industries and tasks. Third, they needed to represent a nationally-recognizable organization and brand. Fourth, traditional organizations for comparison needed to have 15,000 or more employees. These inclusion criteria were required to ensure a large pool of potential, present, and past members who could have reviewed or rated these organizations and be somewhat comparable to the large national organization of the U.S. Military. Fifth, we reasoned that comparison organizations needed to represent a range in terms of their reputation for treating employees well. That inclusion criterion would help ensure that associations found between reviews-and-ratings (if found) were not attributable to a selection-bias. Thus, the Forbes list of America's Best Large Employers for 2019 (Valet, 2019, April 17) was consulted to identify large, traditional corporations that varied in terms of their employee-relations reputations. Ultimately, Michelin, Apple, and Bank of America were identified as meeting all five criteria. Michelin was a top-10 rated company for employee satisfaction in 2018 and 2019. Apple was a top-20 rated company, while Bank of America was not listed in the top-500. Thus, we downloaded the first 100 reviews, which had a minimum of 10 user ratings, for each of the three traditional organizations.

Data analysis
Users of indeed.com can review their organizational experiences on a scale of 1-5, thereby providing a similar measure to the military portrayal variable created in study one. Additionally, users of indeed.com can also rate posted organizational reviews by indicating whether the review was "helpful" or "not helpful." To determine how users of the virtual space valued the reviews that were posted, we added the number of viewers who rated reviews as "helpful" with those who rated the review as "not helpful". The number of helpful ratings were then divided by the total number of ratings to determine the percentage of raters who found the review helpful.

Study 2: results
Descriptive statistics are reported in Table 5. Regression analyses revealed that organizational reviews were significantly and positively predictive of the percentage of raters who found the reviews helpful for all three military branches (Table 5), thus H2 was confirmed and these findings are consistent with study one. In answering RQ1, regression analyses revealed that organizational reviews were significantly negatively predictive of raters' assessment of the helpfulness of reviews for both (b) Bank of America (β ¼ -.43, p < .001) and (c) Michelin (β ¼ -.20, p < .05). There was no significant relationship between organizational reviews and helpfulness rating for (a) Apple. In other words, website visitors tended to find reviews that portrayed the military positively as more helpful, for all three major branches; however, this pattern did not hold for large, traditional corporations in these analyses.

Discussion
The objective of this investigation was to test a theoretical explanation for patterns of organizational reviews and their user ratings on virtual spaces. Namely, using social identity theory and uncertainty management theory, we proposed that when potential, present, and past organizational members' identity and uncertainty needs align, there will tend to be a pattern of rewarding favorable organizational reviews with positive user ratings in virtual spaces. Data from two studies involving reviews of the U.S. Military were consistent with that theoretical explanation. Additionally, we questioned how this pattern of ratings functions in organizations where the identity and uncertainty needs of potential, present, and past members do not necessarily align as consistently. Analysis of the review-and-rating patterns from three large traditional corporations produced mixed results-again, consistent with theorizing. The following paragraphs explain how these studies contribute to the organizational socialization, totalistic organizations, and virtual space literatures.
The rise of online organizational reviews and subsequent user ratings are an increasingly important and influential means by which potential members engage in cybervetting (Berkelaar, 2013), especially given the number of digital natives entering the workforce. Recently, organizational communication scholars have directed attention to the rising influence of virtual spaces within organizational socialization processes (Dailey et al., 2016). Of course, organizations engage in huge efforts to manage their online impressions and reputations, specifically to improve recruiting goals (Avery & Mckay, 2006). However, some third-party virtual spaces-such as the two sites of research for this study-are difficult for organizations to manage directly. That quality makes these sites appealing to potential members who seek out information that is not curated by the organization and deemed more authentic (Piercy & Lee, 2018).
To our knowledge, this study is the first to contribute a theoretical Note: Mean listed; SD in parentheses. * ¼ p < .05, **p < .01, ***p < .001.
explanation for the kinds of patterns that can be expected of third-party, online organizational reviews and ratings. In this way, this paper contributes to the organizational socialization literature the notion that identity and uncertainty needs may explain what to expect from online organizational review sites. Namely, where potential, present, and past members tend to be invested in managing their identity and uncertainty needs-as with the totalistic organization of the U.S. Military-online users will tend to reward favorable organizational reviews with positive ratings. Membership in the U.S. Military is somewhat unique in that professional identification and organizational identification align strongly because a soldier or sailor cannot perform that identity and salient group membership in other organizations (within the U.S.). That exigency means the identity and uncertainty needs of potential, present, and past members will tend to align, resulting in the positive review-andrating pattern. Conversely, those needs tend to be mixed for members of traditional, large corporations-a point identified by the negative or neutral review-and-rating patterns identified in the second study. Future research could extend theorizing by exploring and testing more conditions under which members of traditional organizations reward favorable organizational reviews with positive ratings online. For past members, the profound influence of totalistic organization (TO) socialization on members' identity and values (Hinderaker, 2015;Howe & Hinderaker, 2018) may mean that exit presents an existential dilemma. Absence from the military reduces the former members' ability to perform their identity (Howe & Shpeer, 2019). Thus, alternative forums for communicating that identity would supply former members of TO with enjoyment and a strong sense of place and affirmation (Fieseler, Meckel, & Ranzini, 2014). Previous researchers have found that veterans' well-being is increased by socializing with other veterans in organizations, such as Veterans of Foreign Wars and Disabled American Veterans (Ford, Northrup, & Wiley, 2009). Recent research has addressed the role that technology plays in configuring social networks (Lee & Katz, 2015). Therefore, after the United States' longest wars, and considering the widespread use of modern information communication technologies, it is not surprising that alternative story-swapping forums are becoming increasingly available in virtual spaces.
These findings also contribute to previous investigations of counterinstitutional virtual spaces (Gossett & Kilker, 2006) an investigation of a pro-and a neutral institutional virtual space. Virtual spaces are growing in importance as platforms for members' displaced dissent (Garner, 2017;Kassing, 2011). Virtual spaces can be semi-permanent and widely accessible (Foot, Warnick, & Schneider, 2017), and thus, can allow geographically-and temporally-distributed members and former members to compare their negative workplace experience and organize their resistance efforts. Such efforts are neither deviant nor pro-social necessarily. The important idea is that virtual spaces afford potential, present, and past members a means of both producing and consuming socialization information as well as aiding in making sense of organizational experiences-a means which was not as available even a few decades ago. The semi-permanent and widely accessible nature of virtual spaces means that nonmembers can lurk and consume the unofficial and backstage information posted on these websites. Such online anonymous posts could damage organizational reputation or bolster organizational image among potential, present, and past members. To date, Gossett and Kilker's (2006) case of radioshacksucks.com is perhaps the best known and most complete investigation of a counter-institutional virtual space. The present investigations of americangrit.com and indeed.com contributes to the virtual space literature by investigating virtual spaces that include positive, negative, and neutral reviews that bolster and, at times, challenge organizational image. Future research could continue to explore and explain the patterns of reviews-and-ratings on pro-, counter-, and neutral institutional virtual spaces.

Limitations and conclusion
Similar to all investigations, this work has limitations. First, many details of visitors and posters to the websites remains unknown, thereby reducing the certainty with which some claims can be made about the identity, demographics, and motives of virtual space participants. Ethnographic and interview data could supplement these findings by exploring post-exit message exchange observationally and through retrospective accounts. Second, causal relationships between study variables cannot be conclusively determined in the absence of an experimental design, thus, while the hypotheses were confirmed and supported by theorizing, results should be interpreted with some caution. Future studies could employ experimental design to explore whether positive or negative reports effect participants' story evaluations. Although, the current investigation has the advantage of exploring naturalistic message production, which was not influenced by researchers or via obtrusive measurement.
These studies explored patterns of organizational review-and-ratings in third-party virtual spaces. Social identity theory and uncertainty management theory were used to explain the patterns of review-andratings found on these sites. Importantly, content analyses of two different online sites revealed a consistent pattern: U.S. Military reviews were rewarded by website reviewers with positive ratings when those reviews were favorable. The totalistic and identity-shaping nature of membership in the U.S. Military likely explains that pattern. Conversely, and consistent with this theorizing, that pattern was not found for large, traditional corporations in a second study. These findings suggest that visitors to third-party organizational review sites should be cautious about the kinds of inferences they draw from other users' reviews and ratings. The patterns identified here suggest that users are selectively motivated in their presentation and consumption of organizational reviews.

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.