Abstract
Purpose Social media has created an environment providing readily available information that impacts how individuals seek information and make product decisions. The purpose of this study was to examine adult consumer use of social media for information seeking and the effect on intention to purchase oral health care products.
Methods A cross-sectional study design with a non-probability sample of adult consumers recruited through crowdsourcing with Amazon’s Mechanical Turk (MTurk) was used to examine the effect of social media on purchasing of oral health products. Data collection included demographic, social media use, and the Information Acceptance Model (IACM) which measured information acceptance from friends and customer reviews from social media. Descriptive statistics, subgroup analysis of demographic variables, and multivariable linear regression were used to assess the relationship among variables.
Results Participants (n=308) were mostly males (56.2%), average age 37 years, held bachelor’s degrees (68.8%), from the southern United States, and had been to the dentist within the last two years (72.8%). Instagram (40.6%) was the most frequently used platform and social media was frequently used (36.7%) to obtain information on oral health care products. Findings from the IACM information subscale indicated high scores for information from friends and customer reviews (p<0.01). The information needs (p=0.049) and information adoption from friends (p=0.03) were slightly less for females when compared to males. Friends’ recommendations (p=0.02) on social media had the greatest effect on intention to purchase among participants living in the Midwest. Multivariate regression models indicated information usefulness and information adoption had the greatest influence on purchase intention from friends (p<0.001), while information credibility, information usefulness, and information adoption had the greatest influence from customer reviews (p<0.001).
Conclusion Results provided some initial insights into the way consumers may use social media for information about oral health products and the influences on them choosing to purchase the products. Oral health care providers can use this information to educate and guide patients to credible information sources in their search for oral health care products.
- social media
- purchase intention
- oral health care products
- consumer information seeking
- electronic-word-of-mouth
INTRODUCTION
The availability of enormous amounts of information and communication through social media has impacted all aspects of life. Web 2.0, the most recent internet version, is more user driven and allows for greater creativity and exchange of information, making it part of most everyone’s personal and professional lives.1 This interactive platform has created a means for electronic-word-of-mouth (eWOM) through discussion forums, videos, blogs, and a variety of social media websites that encourage communication and has also become a source of information for consumers.2–5
According to the PEW Research Center, approximately 70% of people use some type of social media site such as Facebook, YouTube, Instagram, Pinterest, LinkedIn, Snapchat, Twitter, WhatsApp, TikTok, Reddit, and Nextdoor in the United States (US).6 Use of social media varies by age group, with a little over 80% of adults aged 18 to 49 using social media sites. The Pew Research Center also indicated social media use decreased with age, with social media use at 73% for ages 50 to 64, and 45% for those over the age of 65.6
Because more people are relying on social media as sources of information, it has changed the way patients or consumers behave, impacting their decision-making processes.4,6-9 In addition, individuals have increasingly been using social media for health information.10-12 Estimates of the public’s use of social media for seeking health information ranges from 72% in the US and 61% globally.13,14 Although little is known about oral health information seeking on social media, research from Saudi Arabia suggests 52% to 68% of consumers have used the internet and social media to search for oral health information.15,16 Social media increasingly fills the need for health information seeking and some consumers may turn to these platforms rather than consulting with health care providers.4,5,10 This is problematic because misinformation on social media is prevalent and some recommendations could potentially cause harm.10,17
With large numbers of individuals accessing social media, companies can produce targeted marketing content to influence consumer behavior without individuals even being aware.2–6 Businesses invested over $108 billion in 2018 in digital marketing (including social media) and expected growth to over $150 billion in 2020, indicating its growing use as a tool for influencing consumer behavior.2-4,6,7 Many of these individuals receiving targeted marketing are patients of oral health care providers and may be receiving information on product choices for oral health care.
Consumers may not be aware that advertisements and information can be targeted to them using algorithms based on their social media contacts, areas of interest from previous searches, and geographic location.18 They may also be unaware that many of the resources they find are sponsored sites that receive payment for recommendations, many of which are paid social media influencers used to market and encourage individuals to try products.19,20 However, social media influencers may also be family members, friends, or colleagues.20 Because of the ease of access to information through social media, oral health care providers are likely to see an increased number of patients influenced by social media when seeking information and choosing products. An extensive search of the literature in academic databases including EBSCO, PubMed, and Scopus conducted with the assistance of an academic librarian did not reveal research specifically related to the effect of marketing through social media on the intention to purchase oral health care products, indicating the need for research in this area.
The purpose of this study was to examine adult consumer use of social media for information seeking and the effect on intention to purchase oral health care products. The first aim was to describe acceptance of information on social media, the second aim was to measure demographic subgroup (age, gender, education, etc.) differences in information acceptance or intention to purchase, and the third aim was to model the predictive relationship of attitude about information and intention to purchase oral health care products.
METHODS
A cross-sectional survey research design was used to gather information from a non-probability sample of adult consumers to examine the use of social media for information seeking and effect on intention to purchase oral health care products. The study received approval from the MCPHS University Institutional Review Board (IRB), protocol number IRB-2021-2022-98.
Sampling procedure
A web-based survey (Qualtrics, Provo, UT, USA) was selected for this study. Many consumers utilize social media to seek information about a potential purchase, making a web-based survey an ideal manner to distribute and gather survey data for this study.15 A non-probability sample of adults in the United States was recruited to participate through crowdsourcing using Amazon Mechanical Turk (MTurk).
Crowdsourcing was developed in 2006 and provides a platform to gather responses from a large group of people via online communities.21 The crowdsourcing platform offers a wide pool of participants streamlining both the recruitment and data collection processes and has been used in psychology, public health, and many other disciplines.21–23 Research has shown crowdsourcing, such as MTurk, can provide a cost-effective way to gather data from national or international samples.21,24 MTurk workers review the task to complete the survey, also referred to as a HIT (Human Intelligence Task), and determine their eligibility and interest in participation.
Inclusion criteria for participants were being at least 18 years of age, residing in the US, able to read English, and able to complete the online survey. Exclusion criteria included individuals that had previously or presently worked or who have close family members that worked in the dental field, were under 18 years of age, unable to read English, or unable to complete an online survey.
A power analysis (G*Power) for the most conservative planned statistical test (chi-square test of independence, two-tailed, df=5) using a medium effect size w=.3, a=.05, and 80% power suggested a minimum sample size of n=143. Adjusting for expected attrition of 30%, the final recommended sample size was n=205.
Instrument
The final survey contained 33 items, divided into nine sections consisting of demographics (7 items), social media use (6 items), information quality (3 items), information credibility (4 items), needs information (2 items), attitudes toward information (3 items), information usefulness (2 items), information adoption (2 items), and purchase intention (4 items).
The validated instrument used for data collection was based on the Information Acceptance Model (IACM). The IACM was developed based on the Information Adoption Model (IAM) and two components of the Theory of Reasoned Action (TRA).3 The TRA is used to predict behavioral intentions meaning the stronger the intention to do a behavior, the more likely the behavior will be performed. The IACM was designed to evaluate the influence of social media information from friends and anonymous sources and the behavior of the consumer based on that information. In this study, the behavior was the intention to purchase an oral health product. The IACM instrument had a seven-factor structure that included the following subscales: information quality (IQ), information credibility (IC), needs of information (NOI), attitudes towards information (ATI), information usefulness (IU), information adoption (IA), and purchase intention (PI).
Convergent validity for the subscales was assessed with composite reliability (CR) and all scales had a value of greater than 0.80 (ranging from 0.815 to 0.854) with anything over .70 being considered acceptable suggesting the IACM subscales all had adequate convergent validity.3 In addition, factor loading was above .70 (ranging from 0.72 to 0.89) for all subscales.3 Discriminant validity showed each subscale was measuring a different construct.3 Structural modeling technique was used to test the IACM as a predictive model and the goodness-of-fit indices indicated the model fit the data from the validation study sample (GFI 0.930, RMSEA 0.047, p<0.001).3 Permission was obtained from the author for use of the IACM instrument.
Procedure
A HIT, or virtual task, was created in Amazon MTurk to describe the purpose of the study and distribute the invitation to only US residents with a link for the survey. The link then took participants to the informed consent. The informed consent asked participants to indicate whether they met the inclusion criteria and provided agreement to participate in the survey.
Participants were compensated 50 cents for survey completion. The MTurk workers’ identity was anonymous and not connected to their survey responses. Response settings in the electronic survey platform did not allow participants to use a back button to change responses. Security settings also prevented indexing on search engines, and anonymized responses by not collecting IP addresses. Once the participants completed the survey, they were issued a unique identifier to submit to MTurk for compensation. Participants could withdraw from the survey at any time. The survey was opened to participants on May 24, 2022, and closed the same day when the target sample size was reached. Some participants were in the process of completing the survey, so the target sample size was exceeded. After survey completion, one of the investigators confirmed the identifiers to approve compensation for the respondents by MTurk.
Data Analysis
Measures of central tendency (mean, median), and variance (standard deviation, Inner Quartile range) were calculated for continuous variables and frequency for categorical variables. Distributions for all continuous variables were analyzed for statistical assumptions including normalcy and co-linearity. No variables were transformed but non-parametric tests were used in each instance where the assumption of a normal distribution was not met. Analysis of data was done to determine the presence of missing values and outliers prior to data analysis and any individual missing more than 20% of their survey responses were removed using listwise deletion. A p<0.05 was used as the alpha cutoff for statistical analysis. All analysis was conducted using a statistical software program (SPSS version 28; IBM Corp., Armonk, NY).
Subgroup analysis of demographic variables was conducted using Spearmen’s Rho when comparing two continuous variables and the Mann-Whitney U rank order of medians test when comparing values between two different groups (e.g., associate and bachelors). Multivariate linear regression was used to model the simultaneous influence of information attitude and adoption on purchase intention. All IACM subscales were entered as predictors plus demographic variables that showed a statistically significant relationship with the purchase intention variable.
RESULTS
Demographics
A total of 320 people opened the survey link. Eleven participants with fewer than 80% complete responses and one outlier were removed from the analysis resulting in the final sample size of 308 and a completion rate of 96%. The median participant age was 37 years with a standard deviation of 10.25. The minimum age of participants was 19, and the maximum was 69 years and over half identified as male (56.2%). The majority of participants were White (85%), came from the southern region of the US (40%) and two-thirds held a bachelor’s degree (68.8%). Most had gone to the dentist within the last two years (72.8%). All demographic information is displayed in Table I.
Respondent characteristics (n=308)
The highest daily use of social media was Instagram (40.6%) followed by Twitter (28.2%) and Facebook (27.9%). A quarter of participants used social media three days per week for a total of 10 hours or less per week (55%) (Table II). About 40% of participants indicated they used social media sometimes to seek information to buy items, while about the same number indicated frequent use of social media to help obtain information on oral health care products and sometimes purchased oral health care products because they had seen them on social media.
Social media type (n=308)
Acceptance and Adoption of Information
Results indicated participants had high or close to maximum scores across all subscales for both friends and customer reviews. However, all pairs of subscales (information quality friends and information quality consumer reviews) had a statistically significant difference and in all cases the subscale associated with friends had a higher value than consumer reviews. The effect size of each test was calculated as r = z/√n, where z is the standardized statistic from the Mann-Whitney U test (n = 308). The resulting r statistic can be interpreted as .10 to <.30 is small, .30 to <.50 medium, and .50 as large. Information credibility, attitudes towards information, and information adoption showed medium effect sizes (.34 to .38) and had the largest effect sizes among subscales. All other subscales had a small effect size with information quality, need for information, and purchase intention as the lowest (.22 to .23). Descriptive statistics of all information subscales are shown in Table III.
Information acceptance subscales (n=308)
Each of the subscales were correlated with the purchase intention subscales. Table IV shows a large correlation for each subscale comparison. Demographic variables were examined to determine relationships with information adoption (IA) subscales. Increase in age was associated with a decrease in the information quality scale (rho=−.12) however this correlation is considered small. All other correlations with age were not statistically significant p>0.05.
Purchase intention correlation (n=308)
A Mann-Whitney U rank order test of medians was used to examine differences in IA scales between gender groups. Needs of information (NOI) from friends for females (M=7.90, SD=1.42) was slightly less than males (M=8.16, SD=1.55), p=0.049 and r=.11. Information adoption from friends was slightly less for females (M=7.82, SD=1.56) when comparing to males (M=8.15, SD=1.60), p=0.03 and r=.13. The state of residence for participants were categorized using the US Census four region model to analyze differences by region. When evaluating regions with information acceptance variables, people in the Midwest had the highest emphasis on purchase intention (PI) from friends’ recommendations (M=17.08, SD=2.30, p=0.02), compared to the west (M=15.93, SD=2.70), south (M=15.60, SD=2.99) or the northeast (M=15.63, SD=2.87). There was no difference between the other regions. No correlations were found with length of time or reason for last dental visit with information acceptance variables, p>0.05.
Influence of Information Attitude and Adoption on Purchase Intention
Two multivariable regression models were used to assess the influence of IA scales on purchase intention. Model 1 used all IA scales associated with friends regressed onto PI related to friends. Model 2 used all customer review IA subscales regressed onto customer review PI. Results for both regression models are shown in Table V.
Multivariable regression models (n=308)
Model 1. Analysis indicates the model was a good fit for the data (Table V). Inspection of the predictors showed information credibility (IC), attitudes towards information (ATI), information usefulness (IU), and IA were all related to PI for friends. After controlling for all other subscales, IU and IA had the greatest influence on PI. A one unit increase on either of the subscales predicts an approximate .40 increase in the PI friends scale sum.
Model 2. Analysis indicates the model was a good fit for the data (Table V). The IC, IU, and IA scales were all related to PI for customer reviews. After controlling for all other subscale sums, IC, IU, and IA were related to a one-unit increase is related to an approximate .44 increase in the PI customer review scale sum.
DISCUSSION
The use of social media allows individuals access to a nearly unlimited amount of information for any purchase they choose to make. There is limited research about adult intention to purchase based on social media information seeking related to oral health products. This explored adult consumer use of social media for information seeking and its effect on intention to purchase oral health care products. Adding to this knowledge base is important due to the impact purchasing oral products has on a person’s oral health. A better understanding of what influences consumers when using social media for information can allow oral health care providers to better guide patients in their search for products and information and help patients understand how corporations use marketing strategies to influence consumers through social media.
Consistent with findings in other studies, respondents who answered the survey items regarding social media and consumer decision making were males in their mid-30’s.8 Pew Research Center results have indicated Facebook as the most widely used social media platform, this was in contrast to findings of this study, which identified Instagram as the most frequently used platform among participants.6 The Pew Research Center also found the majority of respondents used social media on a daily basis, however, the majority of participants in this study reported using social media three days per week.6
Participants in this study indicated that they turned to social media as a source of information for oral health care products even though the majority had been to the dentist within the last two years. While the respondents had opportunities to make inquiries with their oral health care providers, many reported purchasing oral health care products seen on social media.
Wang et al25 found age, gender, ethnicity, and education level had a significant relationship with seeking health related information online, however this study found few significant correlations between demographic variables and information acceptance. Results from this study identified a small correlation with increased age and a decrease in IQ. In addition, results showed NOI and IA from friends were slightly less for females than males, possibly indicating less acceptance of information from friends. An interesting finding was the slightly higher acceptance of information from friends related to purchase intention by individuals living in the Midwest. Information similar to these results could not be found in the literature, which may be because this information was not obtained in previous studies.
This study found information adoption, information credibility, and information usefulness from customer reviews were positively related to purchase intention, consistent with research findings by Leong et al. and Erkan and Evans.26,27 Needs of information had an insignificant relationship in the results, again consistent with Leong et al.26 but inconsistent with findings by Erkan and Evans.27 Perhaps NOI is only impacting purchase intention when the individual pursues the search for information based on a specific need. When comparing information obtained from friends or customer reviews on social media, information usefulness and adoption had similar effects on the two models, however, Erkan and Evans27 found less of an impact from information obtained from friends than anonymous customer reviews. This is consistent with findings that the credibility of information is more important to consumers when seeking information from customer reviews.
Oral health care providers can benefit from an improved understanding of the role social media plays on consumer’s purchase intention based on information obtained from friends and anonymous customer reviews when counseling patients seeking credible information for making decisions about purchasing oral health care products. Further research is needed to assist oral health care providers in understanding how social media is used by patients for oral health information and to identify ways to better reach consumers to provide education and recommendations about oral health care products. Additional research would be beneficial to compare the impact of oral health care provider versus social media recommendations when individuals seek oral health care information and products to purchase. Based on this study’s findings, research is also needed to study the regional influences seen that indicated a high emphasis on friends’ recommendations on social media for purchases.
Limitations
Utilization of a non-probability convenience sample can include the risk of unknown segments of the population not being included due to nonparticipation in crowdsourcing surveys. There is potential for selection bias and the degree to which results may be applicable to the entire population is unknown. When compared with the US Census, the study sample was consistent for the White and Black/African American groups, however females and individuals over 50 years of age were underrepresented. In addition, survey research is self-reported and causality cannot be determined. The actual behavior of participants in terms of purchasing oral health products based on social media recommendations was unknown. Further research is needed reaching a broader range of participants to provide a wider sample base including a more diverse education and gender representation of participants from across the country.
CONCLUSION
Understanding the influence recommendations from both friends and customer reviews on social media have on oral health care product purchases can be beneficial to oral health care providers. Nearly half of participants in this study indicated purchasing oral health care products after seeing them on social media. Significant findings indicated individuals’ perception of credibility of information from customer reviews had an influence on purchase intention. With more products being marketed by companies directly to individual consumers through social media, it is important to understand individual’s perceptions of what they read and where to find credible information for oral health care products. Future research can help provide additional insights to further understand these influences.
ACKNOWLEDGEMENTS
The authors would like to thank Dr. Ismail Erkan, author of the IACM for granting permission to utilize the validated instrument.
Footnotes
NDHRA priority area, Client Level: Oral health care (health promotion, treatments, behaviors, products).
DISCLOSURES
The authors have no conflicts of interest to disclose.
DATA AVAILABLILITY
Data is available upon request to corresponding author.
- Received February 16, 2023.
- Accepted June 11, 2023.
- Copyright © 2023 The American Dental Hygienists’ Association
