How do people really seek information about others?: Information seeking across Internet and traditional communication channels

Abstract

The present research describes the manner in which individuals use various media in the interpersonal information seeking process. Stephens’ (2007) information and communication technology (ICT) succession theory was applied to an interpersonal information seeking context, and hypotheses and research questions about the channels people use to seek information about others of various relationship to the seeker were offered. Two hundred and twenty-five participants responded to a survey about this topic, and they reported a greater likelihood to seek information about less-known targets using channels where they would be unidentifiable. However, participants reported a greater likelihood to seek information about more-known targets using channels where they would be identifiable. Channels such as social networking websites were frequently reported to be useful regardless of whether the target was well known or less known. Properties of these channels and their implications for interpersonal information seeking as well as theoretical implications of these findings are discussed and directions for future research are examined.

Résumé

How do people really seek information about others?: Information seeking across Internet and traditional communication channels

The present research describes the manner in which individuals use various media in the interpersonal information seeking process. Stephens’ (2007) information and communication technology (ICT) succession theory was applied to an interpersonal information seeking context, and hypotheses and research questions about the channels people use to seek information about others of various relationship to the seeker were offered. Two hundred and twenty-five participants responded to a survey about this topic, and they reported a greater likelihood to seek information about less-known targets using channels where they would be unidentifiable. However, participants reported a greater likelihood to seek information about more-known targets using channels where they would be identifiable. Channels such as social networking websites were frequently reported to be useful regardless of whether the target was well known or less known. Properties of these channels and their implications for interpersonal information seeking as well as theoretical implications of these findings are discussed and directions for future research are examined.

Abstract

Wie suchen Menschen nun wirklich Informationen über andere? Informationssuche im Internet und über traditionelle Kommunikationskanäle

Die vorliegende Forschungsarbeit beschreibt die Art und Weise, wie Personen im Prozess der interpersonalen Informationssuche verschiedene Medien nutzen. Stephens’ (2007) Theorie der Informations- und Kommunikationstechnologie-Abfolge wurde auf die interpersonale Informationssuche angewandt. In diesem Rahmen werden Hypothesen und Forschungsfragen formuliert bezüglich der Kanäle, die von Menschen genutzt werden, um Informationen über andere, die in verschiedenartigen Beziehungen zum ihnen selbst stehen, zu suchen. 225 Teilnehmer beantworteten einen Fragebogen zu diesem Thema. Die Befragten gaben an, dass sie eher solche Kanäle nutzen, bei denen sie eher nicht identifizierbar bleiben, wenn sie Informationen über wenig bekannte Zielpersonen suchen. Die Informationssuche bei bekannten Zielpersonen erfolgt eher über Kanäle, bei denen sie identifizierbar sind. Die Befragten gaben außerdem an, dass Kanäle wie soziale Netzwerkseiten sowohl für bekannte als auch weniger bekannte Zielpersonen brauchbar sind. Die Eigenschaften dieser Kanäle und die Auswirkungen für die interpersonale Informationssuche werden ebenso diskutiert wie theoretische Folgerungen der Ergebnisse und Implikationen für zukünftige Forschung.

Resumen

¿Cómo Busca la Gente Información Acerca de Otros?: La Búsqueda de Información a través del Internet y los Canales Tradicionales de Comunicación

La presente investigación describe la manera en la cual los individuos usan varios medios en el proceso de búsqueda de información interpersonal. Aplicamos la teoría de sucesión de información y comunicación tecnológica de Stephens (2007) (ICT) al contexto de la búsqueda de información interpersonal, y las hipótesis y preguntas de investigación sobre los canales que la gente usa para buscar información acerca de otros relacionados con ellos. Doscientos veinticinco participantes respondieron a una encuesta sobre este tema, y reportaron una mayor probabilidad de buscar información sobre metas menos conocidas usando los canales donde serían menos identificados. No obstante, los participantes reportaron mayor probabilidad de buscar información sobre metas más conocidas usando canales donde serían identificados. Los canales tales como sitios de redes sociales de Internet fueron frecuentemente reportados como útiles sin importar si la meta era bien conocida ó no. Examinamos las propiedades de estos canales y sus implicaciones para la búsqueda de información interpersonal. También discutimos las implicancias teóricas de estos resultados y las direcciones para investigaciones futuras.

ZhaiYao

Yo yak

How do people really seek information about others?

Information seeking across Internet and traditional communication sources

Walther and Parks (2002) suggest that some modern relationships have outgrown the theories used to help explain and predict them. As noted by Walther and Parks (2002), theories and approaches such as uncertainty reduction theory (Berger & Calabrese, 1975), social information processing theory (SIPT; Walther, 1992), SIDE model (Reicher, Spears, & Postmes, 1995; Spears & Lea, 1992), and the hyperpersonal perspective (Walther, 1996) all offer unique and valuable insights into interpersonal processes, however, it is important to note that they were not primarily designed to account for the array of possibilities for channel usage in building, maintaining, and ending relationships. The fact that researchers have done little descriptive work in this field underscores a general gap in the CMC literature; although there are a growing number of channels that people can choose to use, little is known about what media choices people make for various interpersonal goals. The current study is designed to be an initial attempt to answer the question of what channels people use to learn information about other people. More specifically, drawing on Stephens’ (2007) information and communication technology (ICT) succession theory, it addresses the question of channel choice used to seek information about a target for a variety of possible relationships between seeker and target. Before discussing this theory, the paper begins by discussing Uncertainty reduction theory (Berger & Calabrese, 1975) as a framework for understanding why people seek information about others.

Uncertainty Reduction Theory

One of the most basic interpersonal goals is reducing uncertainty about other people, which as impression formation, can be seen as a mirroring process to impression management. URT (Berger & Calabrese, 1975) seeks to explain and predict how people form impressions of each other and focuses on the interpersonal communication process in initial interactions, The core assumption of URT is that people’s primary concern when meeting each other is to reduce uncertainty about themselves and each other in order to increase predictability about each other’s (and their own) behaviors. Thus, the theory holds that in situations where an individual is uncertain, that individual attempts to reduce ambiguity about that situation by seeking information pertaining to that specific environment. For example, as a student enrolls in a class they may seek to reduce uncertainty about their professor by seeking information about that person in a variety of ways. By seeking information about their professor their uncertainty about her or him is reduced.

Although URT (Berger & Calabrese, 1975) was originally formulated solely to explain initial interactions, the authors intended for it to be applied to later interactions, and it has been applied to a variety of relationships. This is partly because the need to learn about people extends beyond the initial stage of a relationship. Knobloch and Solomon (2002) argue that uncertainty is not solely part of initial interactions, but that uncertainty management is a maintenance technique in ongoing, close relationships. This may occur in already established relationships because uncertainty is a continuum in a relationship. Planalp and Honeycutt (1985) and Planalp, Rutherford, and Honeycutt (1998) note that there are events that occur within established relationships that increase uncertainty, and these events occur in most relationships at some time. Thus, the need to continue reducing uncertainty about another person does not disappear once you “get to know” that person, but is instead a process of always changing (both increasing and decreasing) levels of uncertainty.

Subsequent research has also shown that URT can be applied to interactions after initial ones. Knapp and Vangelisti (2004) developed the stages of coming together to describe how romantic relationships develop. The second stage, experimenting, occurs after initiating and is dominated by interactive information seeking strategies. Gudykunst, Yang, and Nishida (1985) examined URT across three relationship types (acquaintance, friendship, and romance), and three cultures (Japan, Korea, and the U.S.). They found that URT fits across relationships and cultures, with some variations.

Parks and Adelman (1983) also sought to expand URT to dating relationships. They were especially interested in examining the use of a partner’s social network as an uncertainty reduction source. Consistent with their predictions, greater levels of communication with a romantic partner’s social network was associated with lower levels of uncertainty about one’s romantic partner. This study also noted that sources of information other than interactive dyadic conversation with a target can help reduce uncertainty about said target.

Each of these studies highlight that ongoing uncertainty-reduction needs exist in a variety of relationships, and suggest that different information sources may be useful in addressing these needs. Although URT has guided work on impression formation in multiple contexts such as intercultural interactions (Gudykunst, 1985), and organizational socialization (Miller & Jablin, 1991), research in channel choice for uncertainty reduction is relatively underdeveloped. As Ramirez, Walther, Burgoon, and Sunnafrank (2002) state: “[a]lthough technological advances have produced tools allowing individuals to seek and acquire information… studies specifically investigating their use to seek information about others have yet to surface” (p. 214). The research on uncertainty reduction that does exist generally looks at interactive CMC exchanges (Pratt, Wiseman, Cody, & Wendt, 1999) or compares FtF interactions to CMC interactions (e.g., Tidwell & Walther, 2002; Westerman & Tamborini, 2006). Although these attempts have great value, they do not illuminate the ways people utilize the plethora of channels available to them to learn about others. One approach that can help accomplish this is Stephens’ (2007) information and communication technology (ICT) succession theory.

ICT Succession Theory

Stephens (2007) offers a model of how channel use may occur as a process, rather than focusing on individual channel use separately. As part of this successive model, Stephens suggests that different tasks may be better accomplished by utilizing different patterns of channel usage. The model is articulated for use with organizational tasks, and draws on a list of tasks identified previously by Flanagin and Metzger (2001), as well as an additional task, known as “documentation”. The model offers specific testable propositions about the differential use of channels to accomplish various goals as relationships between people develop.

Although, Stephens’ (2007) model was designed with organizational tasks in mind, its application to interpersonal tasks should be obvious. For example, Stephens posits that mass media will be more effective as a precursor channel in information tasks. If this is applied to the specific information task of learning about other people (reducing uncertainty about them) then we might expect people to use them in learning information about people they do not know well. However, if so-called “text-capable” channels (also known as interpersonal media) are more effective as a successive strategy for informational goals, then we might expect people to use these channels to learn information about others that are better known. That is, individuals become better known at later points in interpersonal relationships (Knapp & Vangelisti, 2004). As such, these interpersonal channels may well prove to be more valuable for more established interpersonal relationships, relationships wherein both parties have more knowledge about one another.

Although the use of traditional mass media may be possible in accomplishing many organizational goals, their use in accomplishing interpersonal goals is less clear. For example, rarely, if ever, will someone be able to use a traditional mass medium (such as television or radio) to learn about another person. However, a hallmark of mass media can be informative to the application of Stephens’ (2007) model to interpersonal tasks. Mass media are anonymous, that is, senders and receivers of messages generally do not know each other. Thus, these channels allow people to use passive, active, (Berger, 1979) or extractive strategies (Ramirez, et al, 2002) because they do not require direct contact with the target. Thus, channels that allow people to learn information about others without that target knowing about the inquiry may behave similarly to mass media. Therefore, channels that allow the use of anonymous strategies should be effective in learning about lesser-known targets, and text-capable channels (interpersonal media that allow for the reduction of uncertainty through text-based interpersonal interaction) should be effective in learning about better-known targets. The first and second hypotheses are based on this logic:

H1: Channels that allow an individual to unidentifiably seek information about a target will be more frequently used when targets are lesser-known by the information seeker than will channels that do not allow for unidentifiable information seeking.

H2: Channels that require a seeker to be identifiable to a target will more frequently used when targets are better-known by the information seeker than will channels that do allow for unidentifiable information seeking.

Although some channels allow for only one type of communication, more recent technological developments have greatly blurred the lines between mass and interpersonal media. For example, online social networking sites (such as Facebook.com or Myspace.com) offer attributes to users of both mass and interpersonal media. Stephens’ (2007) model does not explicitly address these types of channels. However, if mass channels are good for learning about others early in the uncertainty reducing process and “text-capable” channels are good for learning about others later in the uncertainty reducing process, then a channel that combines both would seem to have great utility for many types of relationships. Thus, these channels would be used heavily to learn information about both lesser known and more well known others. Thus, hypothesis three is offered:

H3: Channels that combine mass and interpersonal media qualities will be useful for learning information about both lesser and better-known targets.

Method

Participants

A sample of 225 undergraduate students enrolled in introductory communication courses at a large Midwestern US university responded to a survey in return for course credit. Of these 225 participants, 79 (35.1%) were male, and 141 (62.7 %) were female, with 5 (2.2%) not reporting. Participants ranged in age from 18 to 25 (M = 19.72, SD = 1.59) and their reported race was as follows: 43 (19.1%) Black, 159 (70.7%) White, 9 (4.0%) Asian or Pacific Islander, 7 (3.1%) Hispanic or Latino, 2 (0.9%) American Indian or Alaskan Native, 3 (1.3%) Other, and 2 (0.9%) did not report their race. One hundred and forty-five participants (64.4%) reported being in their first two years of college, while 110 (35.6%) reported being beyond their first two years. Of the 225 respondents, 219 (97.3%) reported that they owned their own computer.

College students were utilized in this study for two reasons. First, they generally have a relatively high degree of technological ability, and thus are presumed to be familiar with the necessary skills to use a wide variety of methods for information seeking. Eighteen to 24-year-olds have grown-up using computers and are the largest demographic to use computers either at home, work, or school (www.census.gov). Second, they are presumably at a stage in their lives where the Internet is likely to be a salient method of information seeking. This seems to be a particularly potent reason because universities generally require students to keep up on their coursework by way of a university provided email account. Additionally, access to the Internet and its wealth of information is relatively easily accessible for most students. Although the ability to seek information on the Internet is not particularly unique to college students, it is certainly a method of information seeking to which they are accustomed. In short, although college students were chosen for this sample in part because of their convenience, they are a reasonable choice because of these aforementioned characteristics.

Procedure

Participants were scheduled in groups of 7 to 25 to report to the test site, complete an informed consent form followed by the survey. Each participant was allotted 30 minutes to complete the survey, although no participant required the full time allotted.

Measurements

The survey posed questions that asked respondents to rate the likelihood of using various communication technologies in situations where they were going to seek information about a variety of potential targets. Participants indicated their likelihood of using particular communication technologies measured by a 5-point Likert-type scale with 1 being very unlikely that they would use certain media to seek information about another person. Higher scores indicated that a specific media type was more likely to be utilized in seeking information about a specific other person. Specific questions followed the form: “How would you go about seeking information about a classmate of yours?” Each participant evaluated information seeking alternatives for each of the following targets: their best friend, a classmate, a family member, a friend, a boyfriend or girlfriend, a potential romantic partner, or a stranger. These targets were chosen in order to address the different choices made in seeking information about people that are less known and more known.

Specifically, respondents rated their likeliness for information seeking using each of the following channels: face-to-face; email; Instant Messenger; asking another person; text message; a phone call; a letter; Friendster, Myspace, or Facebook; or a personal weblog (“blog”). These channels were not explicitly defined for participants. These channels were chosen for several reasons. First, the inclusion of multiple channels was done to better reflect the real life ability for people to use multiple channels for uncertainty reduction goals. Second, these channels were chosen because they represented channels that seemed especially likely to be utilized by the sample chosen. Third, and most importantly, these channels offer an array of possibilities for hypothesis testing. Asking another person about a target and looking at a target’s blog fall into the category of mass channels, in as much as they allow for lower levels of identifiability between seeker and target. Sending an e-mail, IM, or a text message fall into the category of text-capable or interpersonal channels, as do FtF communication with a target, writing a letter and a phone call; for these channels identifiability is almost unavoidable. Social networking sites were included as an option that combines both mass and interpersonal qualities as one could seek information either anonymously, or make oneself identifiable.

Results

To examine the possibility that differences among various relationship types affect the channel individuals will select for information seeking purposes in more detail, classmate, potential romantic partner, and stranger were collapsed into a category called less known targets; best friend, friend, boy/girlfriend and family member were collapsed into a category called well known targets (see Table 2 for means and standard deviations by channel). Additionally, within these relational categories, composite variables were computed for both identifiable (FtF, email, instant messenger, phone, text-message, and letter) unidentifiable (another person, personal weblog) and mixed (social networking site) channels by calculating a mean likelihood of channel use score.

Table 2.  Paired difference t-tests: Less-known vs. well-known targets

Channel Type	Post Hoc Analysis Results
	M (SD)	t	p	η^2
Notes: Negative t-values represent greater likelihood of using a channel for well-known targets. * Indicates significant result after Bonferroni correction.
Face-to-face		−25.28	.001*	.74
Less Known	3.68 (0.63)
Well Known	4.80 (0.37)
Email		−6.31	.001*	.15
Less Known	2.71 (1.03)
Well Known	3.17 (1.25)
Instant Messenger		−6.67	.001*	.17
Less Known	3.40 (0.98)
Well Known	3.83 (1.14)
Another Person		10.99	.001*	.35
Less Known	4.16 (0.74)
Well Known	3.41 (1.02)
Telephone		−27.44	.001*	.77
Less Known	2.87 (0.83)
Well Known	4.56 (0.56)
Letter		−5.27	.001*	.11
Less Known	1.76 (0.79)
Well Known	2.09 (1.10)
Social Networking Website		9.95	.001*	.31
Less Known	3.90 (1.14)
Well Known	3.26 (1.20)
Weblog		4.73	.001*	.09
Less Known	2.58 (1.37)
Well Known	2.35 (1.33)
Text Message		−10.35	.001*	.32
Less Known	2.20 (0.95)
Well Known	2.81 (1.37)

Hypothesis tests Hypothesis one predicted that channels that allow an individual to unidentifiably seek information about a target will be more frequently used when targets are lesser-known by the information seeker than will channels that do not allow for unidentifiable information seeking. A paired-sample t-test indicated that individuals’ reported a greater likelihood of channel use to seek information about a lesser-known target via unidentifiable channels (M = 3.37, SD = 0.81) than identifiable channels (M = 2.77, SD = 0.51), t (224) = -9.78, p < .001, η²= .30. Thus, the data are consistent with hypothesis one.

Hypothesis two predicted that channels that require a seeker to be identifiable to a target will more frequently used when targets are better-known by the information seeker than will channels that do allow for unidentifiable information seeking. A paired-sample t-test indicated that individuals’ reported a greater likelihood of channel use to seek information about a better-known target via identifiable channels (M = 3.54, SD = 0.61) than unidentifiable channels (M = 2.88, SD = 0.88), t (224) = 12.54, p < .001, η²= .41. Thus, the data were also consistent with hypothesis two.

Hypothesis three predicted that channels that combine mass and interpersonal media qualities will be useful for learning information about both lesser and better-known targets. In the present research one such channel—a social networking site—was identified. Two separate one-sample t-tests were used to determine whether participants rated this mixed-identifiability channel to be likely to be used for both well and less-known targets. A test-value of 3 (the midpoint of the channel likelihood of use scale) was used to test whether participants tended to respond in the direction of being “very likely to use” social networking sites at a statistically significant level. Participants responses tended to differ significantly from the midpoint of the scale and in both cases they indicated that they were more likely to use social networking sites to seek information about both unidentifiable targets (M = 3.90, SD = 1.14), t (224) = 11.86, p < .001, η²= .36, and identifiable targets (M = 3.26, SD = 1.20), t (224) = 3.25, p = .001, η²= .05 ². Thus, the data were consistent with hypothesis three.

Post-hoc analyses Although comparing the likelihood of use between identifiable and unidentifiable channels allows for tests of hypotheses derived from Stephens (2007), this procedure does not allow for claims to be made about how likely individual channels are to be used. For example, although a group of channels is statistically significantly more likely to be used than another group, it is possible that one or all channels in that group are not practically likely to be used. In order to analyze the data with regard to the likelihood of individual channel use within relationship type, descriptive statistics for the likelihood of use of each medium across relationship to target were calculated. Means and standard deviations can be found in Table 1. Utilizing a data-analytical method similar to the one Miller, Boster, Roloff, and Seibold (1978) used to examine likelihood-to-use specific compliance gaining strategies in various situations, we report how relationship of target influences the likelihood of media use for information seeking (see Table 3). This data analytic method was chosen because of the descriptive goal of this analysis. Channels were considered to be likely if the mean likelihood of use score was greater than 3.5 and were considered to be unlikely if the mean likelihood of use was less than 2.5. These cutoffs were chosen as decision rules for media choice likelihood because they were the points which fell at plus and minus one standard deviation from the mean likelihood of use score. Mean likelihood of use also fell at approximately the midpoint of the 5-point scale. Thus, our confidence that these cutoffs truly represented an individuals’ inclination or disinclination to use these media was bolstered because the cutoffs fell at opposite sides of the midpoint of the likelihood of use scale.¹

Table 1.  Means (and Standard Deviations) for Information Seeking Channel Choice Across Multiple Relationship Types

Channel	Relationship Type
	Classmate	Best Friend	Friend	Boyfriend/Girlfriend	Potential Romantic Partner	Family Member	Stranger
Notes: Table presents means with standard deviations presented parenthetically. Higher means represent a greater likelihood of using a channel.
FtF	3.07 (1.27)	4.91 (0.35)	4.49 (0.68)	4.93 (.31)	3.98 (0.98)	4.85 (0.53)	1.93 (1.14)
E-mail	3.00 (1.40)	3.03 (1.54)	3.00 (1.43)	3.06 (1.53)	2.71 (1.29)	3.58 (1.51)	2.41 (1.36)
IM	3.43 (1.31)	4.12 (1.32)	4.02 (1.24)	4.13 (1.36)	3.95 (1.20)	3.08 (1.56)	2.66 (1.34)
Another Person	4.02 (1.12)	3.13 (1.41)	3.84 (1.12)	3.43 (1.42)	4.32 (0.93)	3.25 (1.51)	4.13 (1.08)
Text Message	2.03 (1.09)	3.01 (1.64)	2.75 (1.51)	3.26 (1.67)	2.76 (1.44)	2.24 (1.46)	1.66 (0.95)
Phone Call	2.46 (1.23)	4.67 (0.71)	4.14 (0.90)	4.78 (0.58)	4.03 (0.98)	4.63 (0.82)	1.98 (1.15)
Letter	1.77 (1.10)	1.97 (1.26)	1.78 (1.10)	2.12 (1.38)	1.78 (1.10)	2.45 (1.52)	1.73 (1.04)
Social Network Site	3.87 (1.42)	3.62 (1.53)	3.73 (1.35)	3.46 (1.55)	3.91 (1.33)	2.26 (1.48)	3.92 (1.38)
Blog	2.53 (1.52)	2.59 (1.64)	2.47 (1.53)	2.45 (1.58)	2.49 (1.55)	1.88 (1.25)	2.74 (1.65)

Table 3.  Channels Likely and Unlikely to be Used for Information Seeking

Likelihood of Channel Use	Relationship Type
	Classmate	Best Friend	Friend	Boyfriend/Girlfriend	Potential Romantic Partner	Family Member	Stranger
Likely Channels		FtF	FtF	FtF	FtF	FtF
		IM	IM	IM	IM	E-mail
	Another Person		Another Person	Another Person	Another Person		Another Person
		Phone Call	Phone Call	Phone Call	Phone Call	Phone Call
	Facebook, etc.	Facebook, etc.	Facebook, etc.		Facebook, etc.		Facebook, etc.
Unlikely Channels	Phone Call	Letter	Letter	Letter	Letter	Letter	FtF
	Letter		Blog	Blog	Blog	Facebook, etc.	E-mail
	Text Message					Text Message	Text Message
						Blog	Phone Call
							Letter
Likely Channel Means and Standard Deviations	2.95	3.33	2.90	3.32	3.03	3.35	3.03
	(1.03)	(0.73)	(0.61)	(0.61)	(0.61)	(0.69)	(0.94)

Based on the results presented in Table 3, there appear to be three distinct patterns of media use that emerge from these data. In order to categorize the data, sources were selected which were never rated as likely to be used in information seeking regardless of the relationship type. Text messaging, letters, and blogs were never categorized as likely to be useful for information seeking no matter the target. No matter the target, instant messaging and asking another person were not reported as unlikely to be useful. Social networking sites were reported as useful for learning about all types of targets (although boyfriend/girlfriend fell just short of the 3.5 cutoff) except family members, for which they reported as being unlikely to be useful. Finally, people considered phone calls, FtF, and e-mail as either likely or unlikely to be useful depending on the relationship between target and seeker. For this final category, phone calls were unlikely to be used for information seeking about classmates or strangers, but they were more likely to be used for other relationship types. Face-to-face communication was unlikely to be used with a stranger and only slightly more likely to be used with a classmate—otherwise, face-to-face interaction was likely to be used. Email was unlikely to be used to seek information about a stranger but was likely to be used to seek information about or from a family member (see Table 3).

Discussion

Study Findings

This study was designed to be an initial examination of how people use the variety of tools they have at their disposal to seek information about other people. This was an important question to study as previous research has not thoroughly examined this area and it has been recognized as a shortcoming in the literature (Ramirez et al., 2002). This study examined information seeking based on the relationship of the target to the seeker, and made predictions based upon Stephens’ (2007) ICT succession theory. The data were consistent with the three hypotheses offered. This study found that individuals are more likely to use channels which allow them to maintain anonymity to seek information about less known targets. However, when seeking information about more well-known targets, individuals are more likely utilize channels in which they are identifiable as seekers of information. There are also some information seeking tools, such as social networking sites, which offer users the opportunity to be both identifiable and unidentifiable (depending upon how they are used), and are seen as useful for seeking information about both lesser-known and better-known targets.

Although the usefulness of groups of channels was consistent with hypothesized predictions, these predictions tell little about how useful each individual channel actually is. Post-hoc analyses conducted for each channel’s usefulness in seeking information about each relational type to determine how likely each channel is to be used. Interestingly, according to this analysis, there are some channels that seem to lack usefulness for this goal, and others that have wide usefulness.

First, the channels that people generally report as not being useful for seeking information about others no matter the relationship between target and seeker include blogs, letters, and text messaging. It is not surprising that letters are no longer a preferred tool for seeking information, as e-mail provides a faster, more efficient method of sending letters that many people have access to. It is somewhat surprising, however, that a new medium such as personal weblogs would not be used more frequently. Interestingly, although blogs were considered more similar to mass media in that they allow a person to search for information about another person anonymously, other survey results indicate that most bloggers’ intended audiences are friends and family (Lenhart & Fox, 2006). Thus, people may not be reaching their intended audience with blogs. It is also possible that the sample used in this study is not from a population that uses blogs for this purpose. Future research could examine why these channels were not more likely to be used for information seeking about other people.

Second, there are some channels that appear to be useful for social information seeking across relational targets. These include instant messaging and using another person. Social networking sites, such as facebook.com and myspace.com were also generally considered useful, except for seeking information about family members. Consistent with the critical mass theory of interactive media (Markus, 1987) it is not surprising that college students do not find these sites useful for information seeking about family members. These sites are primarily used by high school and college age students—although their popularity has recently been extending to other populations as well (see for example Balakrishna, 2006). Therefore, many of their family members will not have a presence on these sites. Future research should examine why social networking sites are useful and how people form impressions about others in social networking sites.

Theoretical Implications

Of particular interest to the current study is the application of Stephens’ (2007) ICT succession theory to an interpersonal context. Stephens developed ICT succession theory with organizational tasks in mind. However, testing some of the propositions (specifically 3 and 4) in the context of seeking information about other people may highlight the overall usefulness of this approach. It also seems that this approach may have great usefulness in explaining and predicting channel choice for interpersonal communication beyond the current study. Future research in this area may be greatly served by using ICT succession theory as a framework.

Although this study shows the possibility of applying ICT succession theory (Stephens, 2007) to interpersonal communication, this applicability is not without questions. For example, the fifth and sixth propositions suggest that personal and social tasks differ from information tasks. Although this may be true for the organizational task framework for which Stephens suggests the model, it does not necessarily hold in the interpersonal realm the current study applies the theory to. The current study specifically examines the use of different channels for a task that is both personal and informational: learning about someone else. Thus, although some of ICT succession theory’s propositions may apply to interpersonal information seeking, others may need more thought before they can be applied.

It is also interesting to note how these data fit with other contemporary CMC theories. Particularly interesting is how the usefulness of social networking sites for social information seeking, such as facebook.com and myspace.com, challenges these theories. This is a relatively new channel that was not discussed by Walther and Parks (2002) or Ramriez et al. (2002) in their discussion of the Internet’s challenge to traditional impression formation theories.

First, these data suggest narrower boundary conditions for the applicability of the SIDE model to CMC relationships than have previously been suggested (Lea & Spears, 1995). Although SIDE dynamics generally pertain to group-based rather than interpersonal relationships, it has been suggested to drive relational development in some cases. In that application, SIDE predicts that individuals meeting online develop attraction for each other through the interactive discovery of complementary social category memberships. This is distinctively not idiosyncratic, interpersonal information, according to SIDE, but judgments about the social groups to which partners belong. In the present study, however, people reported a high likelihood of using social networking websites to find out information about others, regardless of the relationship with the target (except family members, as noted above). These types of websites generally contain individuating cues about targets, such as photographs, likes and dislikes, comments from other people, etc. Although this does not undermine the usefulness of the SIDE model in some arenas, the fact that people report a high likelihood of using a channel that provides a great deal of visual, physical appearance cues (on which SIDE places a premium) and other idiosyncratic information about a target (as well as groups to which they belong) suggests that people may be able to learn about people as individuals online, and not just rely on social category memberships as predicted by SIDE model. Future research should be conducted to address this possibility, and whether users are drawn to use social networking sites to garner individuating impressions of partners who they have met via systems that offered more categorical than interpersonal information.

Similarly, these data call into question the nature and frequency of hyperpersonal effects online (Walther, 1996). Again, if people are likely to use social networking channels to gain information about others, some dimensions of the hyperpersonal model may be altered with respect to the model’s original application to text-based interaction. Seeing candid pictures and learning more about people based on what others say about them (comments left on their “Wall” such as those found in a social networking site) is the kind of information that is particularly useful for reducing uncertainty (Berger & Douglas, 1981), but seems less likely to facilitate the idealization component of the hyperpersonal model. New questions also arise with regard to the selective self-presentation component. Is selective self-presentation occurring as networkers choose which photos to feature, which hobbies to list, which favorite quotations, movies, and books to list? Allowing people to post comments on their wall may interfere with control over what others see about them, critical to the hyperpersonal framework. On the other hand, those who can post to one’s profile have been selected and enabled, as “friends,” to do so. Perhaps individuals select friends who collude in the construction of desirable self-presentations online (even if that which is desired is conventionally risqué). The nature of friend selection and its possible role in the collaborative presentation of face in social networking sites is a possibility worth exploring. New means of self-presentation deserve examination from this perspective to see if the hyperpersonal model can be extended or if it fails with the use of new communication technologies.

Finally, the popularity of social networking sites questions some notions of SIPT (Walther, 1992) as well. Social information processing theory states that people can form impressions of others online that are as complete as those formed offline, but more time is necessary for the task. One metaphor posits that the broader bandwidth of face-to-face interaction offline with regard to carrying social information for uncertainty reduction is analogous to water flowing through a very wide hose, whereas the relatively narrower bandwidth of social information online is like water flowing through a more narrow hose (Griffin, 2006). In the end, the same amount of water—or social information—can pass through either hose; it simply takes longer through the narrow hose. Over time social information should accrue online the same as it does offline. However, this metaphor assumes that all water passing through the hose is the same sort of water. The high likelihood of using some channels suggests this may not be the case. That is, individuals may find some information more useful than other information in forming or refining impressions about others online. Not only do some forms of online presentation transmit information more quickly—photos versus self-descriptions, for example-- information may be weighted differentially as well. Photos or others’ comments, for instance, might override textual self-disclosures in terms of perceived veracity. Suspiciousness about the truth of online self-presentations guides much online interpersonal information-seeking, and the approaches that information-seekers use to evaluate the veracity of online information are just beginning to be explored (Cornwall & Lundgren, 2001; Ellison, Heino, & Gibbs, 2006). It may be that people are drawn to sites like facebook.com because they provide information that is particularly useful for learning about someone. Simple heuristics, such as judging others’ popularity by the number of friends Facebook purports them to have (Kleck, Reese, Behnken, & Sundar, 2007), or judging their attractiveness based on how attractive their friends are and what their friends say on Facebook (Walther, Van Der Heide, Kim, Westerman, & Tong, 2008) are some contemporary examples. Although SIPT was not specifically designed to address this possibility, future research in this area should examine the variance in information types across channels.

Limitations and Future Directions

As with all studies, the current research has notable limitations to discuss. One centers around the measurement of likelihood. Participants were asked to rate their likelihood of using each channel to gain information about a variety of different people. Although this method is useful, there are other ways this could be measured, as people’s self-reports of predicted behaviors may not accurately reflect their actual behaviors. Thus, instead of using hypothetical likelihood ratings, people could be asked to recall recent experiences centered around actual information seeking situations. Experiments could also be designed asking people to find information about a variety of people and offering them choices, noting which choice they make for each person.

Another possible limitation of the current study is the sample employed. Although there are good reasons to target college students in asking how a variety of channels are used to learn about others, it would be interesting to examine these patterns in other populations. For example, Kim, Kim, Park and Rice (2007) found college students say that the people they communicate with over a variety of channels are close to them. However, this was not true for all others, as there were some channels through which people the people that were communicated with were not as close. If a similar distinction exists for the use of various channels in seeking information about others, the current study is not able to find out, and future research would be useful to do so.

Another limitation of the study involves the groupings of relationship types. The current study compared two groups of relationships: less known compared to more well known people. It is possible that other measures of relationship type could offer even more insight into information seeking behaviors targeting different people. For example, it is possible that relationship type could act as a proxy measure for intimacy. It might be that people need to feel certain closeness to another person before some channels become acceptable to use. The current study has no way to determine this, but future research should ask questions about closeness to target instead of simpler categorical differences only (similar to Kim, Kim, Park, & Rice, 2007).

This investigation is an initial attempt to address the question of how people use the variety of channels available to them to seek information about others. The data suggest that people do use a variety of methods, and that these methods can differ based of the relationship of the seeker to the target. This study points to the need for theoretical refinement for uncertainty reduction across relationships to account the usefulness of the Internet and other channels. It also highlights the need to determine boundary conditions on the major theories in CMC. There is great usefulness in directing future research toward answering these theoretical questions.

Notes

• 1

  Using these points as decision rules for whether media were likely or unlikely to be used was determined to be a relatively stringent decision rule. A series of one-way analyses of variance revealed that by using these decision rules there were substantive differences apparent between the likelihood of using media determined to be likely or unlikely media choices (η² ranged from between .62 and .81)

• 2

  The small effect size for this test may have been due primarily to the inclusion of family members in the category of better-known targets. It is possible that many participants’ family members had little or no presence on social networking sites as these technologies have enjoyed quicker diffusion among younger individuals. In order to test this possibility, an additional one-sample t-test was conducted removing participants’ family members in the well-known target score. The result showed a stronger effect for identifiable targets (M = 3.80, SD = 1.31), t (224) = 6.87, p < .001, η²= .17.