Quantifying social presence

1. Quantifying social presence Li Jin (ljin@kent.edu): Kent State University, Florida State University Albert Ingram (aingram@kent.edu): Kent State University Nov 6th, 2006, 14th Sloan-C International Conference Orlando, Florida

2. Why quantifying social presence? • Social presence is defined as a feature of technology-mediated communication. (Short, Williams, & Christie, 1976) • Social presence addresses the major difference between traditional communication (face2face) and CMC • The need for a quantitative variable dedicated to CMC for in-depth studies

3. Definition of social presence Short, Williams, & Christie (1976): • A quality of the medium • The degree of salience • The interpersonal relationship • Media vary in their degree of SP

4. Definition of social presence A quality of the medium The degree of salience The interpersonal relationship Media vary in their degree of SP A quality of user(users) Measurement of performance Responsive to Inter -personal Relationship Responsive to media

5. Problems with measuring SP • Short, et al (1976): users’ perception of the quality, which will affect their choice of the medium and the interaction. • Walther (1992): users’ feeling about others in the mediated interaction. • Gunawardena and Zittle (1997): the degree to which a user (in CMC) is perceived as a real person • Picciano (2002):a student’s (CMC user’s) sense of being and belonging in a course • --SP is a quality associated with users, responsive to medium’s quality, and other variables. When medium is the controlled variable, SP of individual user varies.

6. Solution: Measure user’s performance • Walther (1992): Critique:…(researchers) do assessments “by asking subjects to rate the media, rather than by assessing their performance in using them” (p. 55) • Garrison, Anderson and Archer (2000): define social presence as “the ability of participants in a community of inquiry to project themselves socially and emotionally, as ‘real’ people (i. e., their full personality), through the medium of communication being used” (p13). • ---We measure individual’s ability by measuring their performance.

7. The evolution of definition of Social presence A quality of the medium The degree of salience Interpersonal relationship Media vary in their degree of SP A quality of user(users) measurement of performance Responsive to Interpersonal relationship Responsive to media

8. The definition of Social Presence social presence is an individual’s (or a group’s) performance to project him/her self (or themselves) as a real person (s) in a specific mediated communication. --We are going to measure User’s performance in CMC.

9. Factors affecting degree of Social presence • Technology used • Content delivered • Instructional methods • Other participants’ performance • Individual user (users)’s status • …

10. The content analysis approach The interaction in Synchronous CMC • Data: Text-based transcription of “conversation” • Lack of audio or visual clues, or textized audio visual clues (Emoticons, sound effects, etc.) • Conversational style: spontaneous, responsive, interpersonal, less lengthy, • Each participation as one unit of analysis. An utterance, an audio effect, an e-icon, etc.

11. The categories Bales’ (1950) categories for studying group interaction 1,shows solidarity, raises other’s status, gives help, reward 2, Shows tension release, jokes, laughs, shows satisfaction 3, agrees, shows passive acceptance, understands concurs, complies 4, gives suggestion, direction, implying autonomy for other 5, Gives opinion, evaluation, analysis, expresses feeling, wish 6, gives orientation, information, repeats, clarifies, confirms 7, asks for orientation, information, repetition, confirmation 8, asks for opinion, evaluation, analysis, expression of feelings 9,asks for suggestion, direction, possible ways of action 10, disagrees, shows passive rejection, formality, withholds help 11, shows tension, asks for help, withdraws out of field 12, shows antagonism, deflates others status, defends or asserts self 1-3:Showing positive reaction 4-6:problem Solving attempts 7-9:asking questions 10-12:Showing Negative reaction

12. The categories Bales’ (1950) categories for studying group interaction 1-3:Showing positive reaction response 4-6:problem Solving attempts 7-9:asking questions initiation Categories (resp. vs. ini.) are exclusive to each other 10-12:Showing Negative reaction

13. The categories Flanders (1970) and his colleagues divided classroom conversation into three large categories: 1.teacher behavior 2.student behavior, and 3.silence/confusion. They suggested that these three conditions “exhaust all the possibilities” (p. 33). Categories 1,2,&3 are inclusive.

14. Quantifying Social presence1:Categorizing Step 1. Identify the data set to be categorized. 20 students and one instructor in our class, there are 13 classes, once each week, for semester Fall, 2006. -participant i,i=1,2,…20. -class j, j=1,2,…13.

15. Quantifying Social presence1:Categorizing Step 2. Filtering non-contributors: -Impolite or negative utterances; -Utterances that are not intended to interact with other participants; -Grammatical supplement of a previous message: Examples: 18:26:34 student 7: people work together 18:26:38 student 7: ? 18:39:30 student 7: do you mean happer? 18:39:33 student 7: hamper

16. Quantifying Social presence1:Categorizing Step 3: Initiation vs. Response A response, as long as it has not been identified as a non-contributor in Step 2, it is a contributor to Social Presence.

17. Quantifying Social presence1:Categorizing Step 4: Questions vs. non-questions For an initiation, if it is a question, normally it is intended to interact with other participants, unless it has been sorted as a non-contributor in Step 2.

18. Quantifying Social presence1:Categorizing Step 5: Social/emotional expressions Initiated utterances that fall into “non-question” category: if it contains any social or emotional expressions, it is count as a contributor. Step 6: others: Those utterances that can’t fit any of the above categories.

19. Quantifying Social presence1:Categorizing Categorizing steps

20. Quantifying Social presence2:Caculating Scores we have for each student i in class j To: total number of contributor Nij1: number of non-contributor Rij: response Iqij: initiated question Ieij: initiated message containing social/emotional expressions (non-questions) Nij2: others

21. Quantifying Social presence2:Caculating Student i’s level of Social Presence in class j: SPij(raw)=Rij+Iqij+Ieij i: the ith participant in each class, i=1,2…n (n=20) j: the jth class, j=1,2,…m (m=13)

22. Quantifying Social presence2:Caculating Standardizing SP raw score: SPi/j(z) : the standardized score of SPi/j raw score; Sj: the standard deviation of SPij(raw) in class j; m: the number of participants in class j.

23. Applications and discussion The meaning of the mean of SP(raw) of each class a. comparisons of means of SP(raw) provide information on the overall tendency of SP b. the instructional method affects the SP overall level: lecture vs. group discussion c. 1st vs. 2nd class: the technology’s effect

24. Applications and discussion Comparison of mean of SP of 13 classes

25. Applications and discussion Normally distributed, with possible outliers around +3 The distribution of SP(z):

26. Applications and discussion Individual’s level of social presence

27. Applications and discussion Use SP(z) as a continuous independent variable, we can explore other variables of interest in different contexts: --Technologies applied in different setting: chat room A and chat room B --Instructional method A and method B --Content delivered --others

28. Applications and discussion High level of Reliability: a.Decisions on contributor vs. non-contributor involve less subjective interpretation: Impolite or negative utterances: common sense/moral standards Utterances that are not intended to interact with other participants: talk to oneself Grammatical supplement of a previous message b.Categorizing process base on observable, non-interpretive characteristics of the message initiation vs. response; question vs. answer; Social/emotional expressions—might involve subjective decisions

29. Applications and discussion Discussions of non-contributors and “others” categories are necessary for interpreting instrument’s validity and reliability 1.Impolite or negative utterances 2.Utterances that are not intended to interact with other participants 3.Grammatical supplement of a previous message 4. “others”: --descriptions --percentages --discussions

30. Quantifying social presence Questions? Comments? Li Jin (ljin@kent.edu): Kent State University, Florida State University Albert Ingram (aingram@kent.edu): Kent State University