581515 research-article2015

HPQ0010.1177/1359105315581515Journal of Health PsychologyThompson et al.

Article

Understanding how adolescents and young adults with cancer talk about needs in online and face-to-face support groups

Journal of Health Psychology 1­–11 © The Author(s) 2015 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/1359105315581515 hpq.sagepub.com

Charee M Thompson1, Brittani Crook2, Brad Love2, Catherine Fiona Macpherson3 and Rebecca Johnson4

Abstract We compared adolescent and young adult cancer patient and survivor language between mediated and face-to-face support communities in order to understand how the use of certain words frame conversations about family, friends, health, work, achievement, and leisure. We analyzed transcripts from an online discussion board (N = 360) and face-to-face support group (N = 569) for adolescent and young adults using Linguistic Inquiry and Word Count, a word-based computerized text analysis software that counts the frequency of words and word stems. There were significant differences between the online and face-to-face support groups in terms of content (e.g. friends, health) and style words (e.g. verb tense, negative emotion, and cognitive process).

Keywords adolescents, cancer, Internet, social support, verbal communication

Background Within the last decade, more attention is being paid to the psychosocial (i.e. psychological and social) aspects of the cancer experience of adolescent and young adults (AYAs) 15–39 year of age, not only assessing AYA mental health outcomes (e.g. Tai et al., 2012), but also identifying AYA support needs (e.g. Goldfarb and Casillas, 2014; Kent et al., 2013). More than 50 percent of AYAs report their needs go unmet for nearly every one of their needs, including a desire for more information about the cancer itself and related health issues (e.g. infertility, nutrition), as well as counseling support (e.g. mental health, sexuality, and intimacy) and family services, including adoption and child

care (Goldfarb and Casillas, 2014; Zebrack, 2009; Zebrack et al., 2014). Unmet needs have been linked to distress for AYAs post-diagnosis (Zebrack et al., 2014), thus underscoring the importance of delivering AYAs appropriate and effective support. 1Ohio

University, USA University of Texas at Austin, USA 3Children’s Hospital Los Angeles, USA 4Mary Bridge Hospital/Multicare, USA 2The

Corresponding author: Charee M Thompson, School of Communication Studies, Ohio University, 43 West Union, Athens, OH 45701, USA. Email: [email protected]

Downloaded from hpq.sagepub.com at UNIV MASSACHUSETTS AMHERST on June 4, 2016

2

Journal of Health Psychology 

Though the challenges and needs of AYAs have become better understood (Kent et al., 2013; Zebrack, 2011), questions still remain about if, and how, AYAs go about seeking and receiving support (Treadgold and Kuperberg, 2010). To contribute to current understandings of AYA needs, we approach this study in two novel ways. First, we explore how AYAs talk about their needs through an analysis of language. In doing so, we provide a more nuanced picture of how AYAs exchange support (i.e. how they talk about their own and other’s needs), as well shed light on what it is about their needs that is important, including how AYAs are experiencing and coping with unmet needs. This is because examining AYA language allows us to gain insight into their “emotional and cognitive worlds” (Pennebaker et al., 2007), as analyses of word use can tap psychological processes that may not be obvious to the individual speaking or to the researcher who is simply coding words for content (Tausczik and Pennebaker, 2010). According to Pennebaker et al. (Chung & Pennebaker, 2007; Tausczik and Pennebaker, 2010), the language individuals use to talk about their experiences gives us information about the focus of their thoughts (topics), including their placement in time (past, present, or future), how they are feeling about situations (e.g. emotions), and how they are processing events (e.g. reappraisal, uncertainty). For example, what can we learn from how AYAs talk about sex and intimacy? Do AYAs use uncertain language, suggesting a need for more information? Do they use great negative emotion words, suggesting a struggle and need for emotional support? Or, do they use more causal language, suggesting attempts to make sense of their experiences? Answering these questions related to AYA language use can inform research and programming by suggesting the types of support (e.g. information, emotional support, tangible services) that might best serve them. Second, our approach departs from those of previous studies because we also address questions as to why AYAs might choose to seek support in either face-to-face or online venues.

According to our knowledge, this study is the first to simultaneously consider AYA use of both online and face-to-face support groups. This is important because both settings are beneficial for AYAs, and yet we know very little about how they may serve different functions. Young adults with cancer may turn to online support communities because they are “tech savvy” and not constrained by the cost, physical location, and meeting times of traditional faceto-face support communities (Bambina, 2007). Moreover, online groups can be accessed whenever it is convenient for AYAs, and AYAs can craft messages at their own pace (White and Dorman, 2001). Online communities also allow anonymity, which promotes honest and intimate self-disclosure among its members (Galinsky et al., 1997). Likewise, the benefits of AYA face-to-face support groups continue to be well-documented (Zebrack and Isaacson, 2012). Whether they are hospital-based support groups, social networking events, sporting tournaments, or summer camps, these programs provide AYAs with opportunities to create friendships and communities that provide emotional support and a level of understanding that AYAs do not feel they can receive from peers without cancer experience (Cassano et al., 2008). Some research also suggests that when given a choice, patients feel more positive about, and perceive more benefits from, faceto-face support than online support (Van UdenKraan et al., 2011). To summarize, we seek to determine how talk about needs differs online versus face-toface. If different emotional and cognitive processes, as suggested by the use of certain language, are occurring in online versus faceto-face contexts, patient support services can be better tailored and directed based on the relevance and function of each medium. We also seek to understand with greater specificity the support needs of AYAs by describing how they talk about their needs. This process involves assessing the relationships between the use of content or topic words (e.g. health, work, family) and the use of style words (i.e. verb tense, emotional process, and cognitive process

Downloaded from hpq.sagepub.com at UNIV MASSACHUSETTS AMHERST on June 4, 2016

3

Thompson et al. words) in face-to-face and online support groups. Thus, we are guided by the following research questions: RQ1: What differences exist in language use between online and offline support groups? RQ2: How do AYAs talk about certain topics, evidenced by relationships between content words and style words?

Methods Sample Data came from two sources. First, full discussion board threads (i.e. original message and all its following comments) were randomly chosen from an archive (years 2007–2010) of a large online AYA cancer support group (6000-plus users age 15–39 years) that is available to AYAs affected by any type of cancer at any stage in their cancer experience (e.g. diagnosis, treatment, remission.). We used a random number generator to select discussion board threads, resulting in a sample of 26 threads or 90 pages of text and 44,233 words. The second source of data consisted of transcripts from five different face-to-face, professionally facilitated cancer support group meetings for a group of eight AYA patients at a large hospital in the Pacific Northwest United States. The oldest AYA in the face-to-face group was in his or her mid-20s. Each meeting lasted at least one and one half hours. Together, the five meeting transcripts totaled 66,997 words. We refer to these two groups hereafter as the “online AYAs” and the “face-to-face AYAs,” respectively.

Procedures Unit of analysis.  For the online AYAs, the unit of analysis was the individual message posting. A posting was defined as what one person says, or the total content, regardless of the length, be it a single word or multiple paragraphs, of a single message submitted to the forum thread (Campbell-Eichhorn, 2008). For the online

group, N = 360 individual posts. For the face-toface AYAs, the unit of analysis was an individual’s conversational turn. A talking turn was defined as all of the words an individual said before, as noted in the transcripts, another individual began speaking (Sacks et al., 1974). The face-to-face group had a sample size of N = 569 conversational turns. Linguistic Inquiry and Word Count. Linguistic Inquiry and Word Count (LIWC) (Pennebaker et al., 2007) is a word-based computerized text analysis software that counts the frequency of words and word stems to study the emotional, cognitive, structural, and process components present in written text or speech (Lieberman, 2008). We included the following topical word categories: family, friend, body, health, sexual, work, achievement, and leisure because previous research analyzing the types of talk in AYA cancer communities (e.g. Love et al., 2012) suggests a wide range of topics and concerns are expressed. The style words we included in this analysis are those related to emotional (e.g. positive and negative emotions, and anxiety, anger, and sadness) and cognitive processes (e.g. general use of cognitive words, and insight, casual, discrepancy, tentative, certain, inhibition, inclusive, and exclusive words) to get a sense of how AYAs experience and think about these topics. Examples of words for each category appear in Table 1. LIWC calculates ratios, or percentages of word categories, as the number of category words in a particular post or conversational turn, divided by total words in that post or turn. Percentages of words, as compared to word counts, were used in analysis to account for individual differences in the amount of content users posted (Pennebaker et al., 2007). Mean ratios for both topic (content) and style words are given in Table 1.

Results As a preliminary analysis, we conducted independent samples t-tests to compare words per unit of analysis (e.g. unique posts and conversational turns) and words per sentence. We

Downloaded from hpq.sagepub.com at UNIV MASSACHUSETTS AMHERST on June 4, 2016

4

Journal of Health Psychology 

Table 1.  LIWC word category examples, mean word frequencies, and MANCOVA results for comparing online and face-to-face support groups. Word category

Example words

Mean frequency Online Group (SD) (N = 360)

Content (Topics)  Family Aunt, bros, family, hubby, .25 (.63) grandkid, parent  Friends Buddy, companion, amigo, .65 (1.75) neighbor, pal, roommate  Body Bone, palm, pulse, chest, 1.08 (1.69) sensation, itch, toes  Health Fever, thyroid, ulcer, 2.49 (3.51) swollen, symptom  Sexual Condom, erectile, breast, 1.93 (4.65) kiss, naked  Work Meeting, taxes, cubicle, .85 (1.23) company, promotion  Achievement Lose, rank, reward, .98 (1.30) confidence, determined  Leisure Cook, run, shop, cocktail, .40 (.88) relax, weekend Verb tense   Past tense Can’t, arrived, became, 3.88 (3.46) called, came, didn’t   Present tense Let’s, miss, I’ve, are, see, 9.11 (4.94) want, has, use   Future tense Will, ought, shall, must, 2.50 (4.06) he’ll, would, Affective processes  Positive Glad, relief, pride, fun, 3.42 (4.23) emotion special, happy, safe  Negative Grim, pain, tough, crap, 1.92 (3.68) emotion tears, battle, terrified  Anxiety Pressure, risk, nervous, .41 (1.26) panic, insecure  Anger Enemy, fiery, resent, 2.14 (4.87) destroy, evil, fight  Sadness Loser, miss, helpless, .67 (1.70) weep, sorrow, ruin Cognitive Unsure, explain, consider, 14.23 (7.77) processes idea, accept  Insight Grasp, understand, found, 2.30 (2.79) info, solve  Causation Make, react, since, affect, 1.72 (2.21) change, control  Discrepancy Besides, must, could’ve, 1.53 (1.84) impossible, lack, hope  Tentative Bet, barely, sort of, 3.04 (2.96) mainly, supposed, seem

Mean frequency F(1, 925) Face-to-face Group (SD) (N = 569)

Partial η2

.37 (2.26)

.89

.00

.30 (2.57)

10.05**

.01

.81 (3.41)

1.61

.00

1.58 (6.84)

3.77a

.00

.18 (1.44)

112.95*** .11

.64 (3.25)

.49

.00

1.12 (4.32)

.80

.00

.68 (3.43)

2.60

.00

9.32**

.01

4.90 (7.31) 12.00 (13.07)

15.71*** .02

.78 (3.06)

78.09*** .08

8.39 (20.28)

18.71*** .02

1.58 (6.23)

.48

.00

.42 (2.28)

.05

.00

.17 (1.09) .35 (2.18) 18.05 (15.54)

130.01*** .12 8.21**

.01

26.99*** .03

2.93 (5.90)

4.20*

.01

1.15 (2.93)

9.73**

.01

1.28 (3.29)

.77

.00

3.14 (6.23)

.38

.00 (Continued)

Downloaded from hpq.sagepub.com at UNIV MASSACHUSETTS AMHERST on June 4, 2016

5

Thompson et al. Table 1. (Continued) Word category

Example words

 Certainty

Absolute, definitely, commitment, clear, all Banned, barrier, avoid, stubborn, stop, wary Add, along, both, close, we, come, with But, either, vs, rather, or, not, just, if

 Inhibition  Inclusive  Exclusive

Mean frequency Online Group (SD) (N = 360)

Mean frequency F(1, 925) Face-to-face Group (SD) (N = 569)

Partial η2

1.56 (1.90)

1.75 (8.95)

.06

.00

.62 (1.89)

.31 (1.37)

10.65**

.01

3.64 (3.13)

4.53 (6.99)

7.96**

.01

2.59 (2.37)

3.60 (6.76)

9.12**

.01

LIWC: Linguistic Inquiry and Word Count; MANCOVA: multivariate analysis of covariance; SD: standard deviation. Means and standard deviations are percentages of total words within each group. *p 

Understanding how adolescents and young adults with cancer talk about needs in online and face-to-face support groups.

We compared adolescent and young adult cancer patient and survivor language between mediated and face-to-face support communities in order to understa...
707KB Sizes 0 Downloads 5 Views