Abstract
The practice of unsolicited advertisements on social media has grown prevalent. This data article presents 837 US-based social media users’ consumer perceptions of such advertisements. Understanding how consumers perceive unsolicited advertising is vital to developing effective digital marketing strategies. Data collection was via an online survey adopting multi-measurement items from extant studies for reliability and validity. The data showed high internal consistency with Cronbach’s alpha testing, and confirmatory factor analysis (CFA) found the measurement model valid. Goodness-of-fit indices showed a good fit with the data. Finally, convergent and discriminant validity was confirmed using the composite reliability, average variance extracted (AVE), and correlations among constructs. Further research may utilise the data using inferential analysis techniques to add to our understanding of consumer perceptions of unsolicited advertising on social media.
Dataset: 10.17632/zcp5httw5c.1.
Dataset License: CC BY 4.0
1. Summary
As digital technology has rapidly developed, social media has become one of the most popular ways for firms to reach a large number of target consumers with their advertisements. However, the practice of unsolicited social media advertisements has grown prevalent with technological innovations. Therefore, understanding how consumers perceive unsolicited advertising is vital for firms to develop effective digital marketing strategies. In this regard, data in this article provides information on how individuals in the US perceive unsolicited social media advertisements based on the constructs such as ad avoidance behaviour, feelings of intrusiveness, perceived ad benefits, and privacy concern. An online survey was distributed to US-based social media users for data collection, and as a result, a total of 837 samples were compiled. All the multi-measurement items for the main constructs were adopted from the extant studies, to support their reliability and validity. A descriptive analysis of the main constructs was provided to understand the data better. In addition, the dataset was rigorously tested for validity and reliability for the re-use in further scientific and scholarly research. Cronbach’s alpha analysis showed that the multi-measurement items for each construct have high internal consistency. Then, confirmatory factor analysis (CFA) was conducted to test the validity of the measurement model. Goodness-of-fit indices showed that the measurement model demonstrated a good fit with the data. Additionally, our data’s convergent and discriminant validity were confirmed using the composite reliability, average variance extracted (AVE), and correlations among constructs. Thus, future researchers may employ inferential analysis techniques utilising the data to provide a deeper understanding of consumer perceptions towards unsolicited advertisement on social media.
2. Data Description
An online survey was distributed to US-based social media users, and as a result, a total of 837 respondents participated in the survey. Table 1 shows detailed demographic information of the respondents on gender (Male: 55.2%, Female: 44.8%) and age (M = 37.79, S.D. = 11.97). Table 2 shows the frequency of social media use of the participants. Overall, 78.1% of the respondents answered that they use social media every day, and 16.4% of the respondents use social media a few days a week. Only 4.5% and 1% of respondents use social media a few days a month and once a month or less, respectively.
Table 1.
Demographic profile of participants (n = 837).
Table 2.
Frequency of social media use (n = 837).
Table 3 shows detailed measurement items for all the constructs in the data with Cronbach’s alpha coefficients. All the constructs in the data were measured with multiple measurement items to better capture the subjective properties of the constructs. Avoidance behaviour and perceived benefits for social media advertisements have four measurement items each. Privacy concern and feelings of intrusiveness have six measurement items each. As shown in the table, all Cronbach’s alpha coefficients are greater than 0.80, suggesting high internal consistency of the measurement items for each construct.
Table 3.
Measurement items and Cronbach’s alpha for constructs.
The descriptive statistics for all the constructs in the data are presented in Table 4. It shows the abbreviated form of all the measurement items in the dataset, the minimum and maximum value, and the mean and standard deviation of the measurement items.
Table 4.
Descriptive statistics of constructs.
Confirming the reliability and validity of data is critical to conduct further inferential analysis and make the best use of the data. Firstly, the validity of the measurement model was assessed with confirmatory factor analysis (CFA) using AMOS. The results of goodness-of-fit indices (x2/df = 3.760, p < 0.01, IFI = 0.966, NFI = 0.954, CFI = 0.966, and RMESA = 0.057) showed that the measurement model demonstrated a good fit with the data [1,2]. Secondly, the convergent validity of the data was tested based on the criteria recommended by Fornell and Larcker [3]. The test results showed that the factor loadings of all measurement items of each construct were greater than 0.70, and they were all significant (p < 0.001), as shown in Table 5. In addition, as shown in Table 6, the composite reliability of each construct exceeded 0.80, and the average variance extracted (AVE) for each construct exceeded 0.50. These results suggested that all the conditions for the convergent validity of the data were met. Lastly, the discriminant validity was tested by using AVE and the correlations between constructs. As shown in Table 6, the lowest value of the square root of AVE (0.788) exceeded the highest bivariate correlation (0.760). This result confirmed the discriminant validity of our data [3]. In sum, all the results of the reliability and validity tests confirmed the adequacy of our data for further inferential analysis.
Table 5.
Standardized factor loadings of measurement items.
Table 6.
Correlations with composite reliability and AVE.
3. Methods
The data were collected using an online survey method in December 2021 as part of a research project on understanding consumer perceptions towards unsolicited advertisements on social media. The respondents were first asked two screening questions—‘Do you currently use or have you used any social media?’ and ‘Have you come across any unsolicited advertisements when you use social media?’—which aimed to ensure that all the respondents were social media users and that they understood what unsolicited social media advertisements were. The screening questions were followed by a question on their social media use frequency. The respondents were then asked the extent of their agreement to the main constructs on a seven-point Likert scale anchored from strongly disagree to strongly agree.
All the measurement items for the main constructs were adopted from extant studies, to support their reliability and validity. The measurement items for advertisement avoidance behaviour were adopted and modified from Cho and Cheon [4] to qualify in the context of social media advertisement. The measurement items for perceived advertisement benefits were adopted from Bleier and Eisenbeiss [5]. Privacy concern and feelings of intrusiveness were measured using measurement items modified from Dolnicar and Jordaan [6] and Edwards et al. [7], respectively.
The online questionnaire was developed using Qualtrics and distributed to US-based consumers through MTurk. MTurk is a crowdsourcing marketplace that offers researchers access to a diverse, on-demand survey panel. Researchers can access a large number of registered panels by offering small monetary incentives. Since MTurk has often been used for data collection, previous studies attempted to confirm the credibility of MTurk as a data source for academic research purposes. In this regard, Buhrmester et al. [8] and Holden et al. [9] found that data collected through MTurk are reliable and have strong test-retest reliability.
A total of 837 data samples were compiled from the online survey. The final dataset was coded in SPSS so that we could conduct the initial descriptive analysis. Subsequently, both SPSS and AMOS were used to assess the reliability and validity of our data.
4. User Notes
The information collected measures perceptions on unsolicited social media advertisements, privacy concern, a feeling of intrusiveness, and ad benefits. Numerous researchers are recognising the importance of the topic, and as such, the data provide a valuable reference for future research to produce further insights into the subject. The dataset is rich, with a sampled population of 837 social media users based in the US. Researchers and industry practitioners can benefit from inferential analysis of the collected data, which can be utilised with confidence in the information’s integrity as the dataset was rigorously tested for its validity and reliability. For example, structural equation modelling and regression analysis could be adopted to analyse the dataset, to understand the relationships between constructs with specific directions of influences. In addition, ANOVA and a t-test could be adopted to compare the perceptions towards unsolicited social media ads across different gender and age groups. The results of the inferential analysis on the dataset will be helpful in conceptualising, designing, testing, and executing more effective social media advertisement campaigns. However, the dataset is not without limitations for users. These cross-sectional data were collected by using a survey method, which makes it difficult for data users to conclude a causal relationship between the constructs. In addition, the dataset was collected in the US alone; additional data and analysis might be required to generalize the results for consumers with different cultural backgrounds.
Author Contributions
Conceptualization, R.R.; methodology, R.R.; project administration, R.R.; writing—original draft preparation, R.R. and J.H.; writing—reviewing and editing, R.R. and J.H.; investigation, J.H.; formal analysis, J.H. All authors have read and agreed to the published version of the manuscript.
Funding
This research had no external funding.
Institutional Review Board Statement
The ethics committee of Middlesex University Dubai approved the data collection before implementation. The authors addressed all ethical concerns during the data-gathering process. Participation was entirely voluntary, and the respondents were informed that the collected information would be used only for academic purposes and remain confidential.
Informed Consent Statement
Informed consent was obtained from all subjects involved in the study.
Data Availability Statement
The data presented in the descriptor are available at doi:10.17632/zcp5httw5c.1.
Conflicts of Interest
The authors declare no known competing financial interest or personal relationships that could have or appeared to influence the work reported in this paper.
References
- Bagozzi, R.P.; Yi, Y. On the evaluation of structural equation models. J. Acad. Mark. Sci. 1988, 16, 74–94. [Google Scholar] [CrossRef]
- Hu, L.; Bentler, P.M. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Struct. Equ. Model. 1999, 6, 1–55. [Google Scholar] [CrossRef]
- Fornell, C.; Larcker, D.F. Evaluating structural equation models with unobservable variables and measurement error. J. Mark. Res. 1981, 18, 39–50. [Google Scholar] [CrossRef]
- Cho, C.; Cheon, H. Why do people avoid advertising on the Internet? J. Advert. 2004, 33, 89–97. [Google Scholar] [CrossRef]
- Bleier, A.; Eisenbeiss, M. The importance of trust for personalized online advertising. J. Retail. 2015, 91, 390–409. [Google Scholar] [CrossRef]
- Dolnicar, S.; Jordaan, Y. A market-oriented approach to responsibly managing information privacy concerns in direct marketing. J. Advert. 2007, 36, 123–149. [Google Scholar] [CrossRef]
- Edwards, S.M.; Li, H.; Lee, J.H. Forced exposure and psychological reactance: Antecedents and consequences of the perceived intrusiveness of pop-up ads. J. Advert. 2002, 31, 83–95. [Google Scholar] [CrossRef]
- Buhrmester, M.; Kwang, T.; Gosling, S.D. Amazon’s mechanical Turk: A new source of inexpensive, yet high-quality data. In Methodological Issues and Strategies in Clinical Research; Kazdin, A.E., Ed.; American Psychological Association: Worcester, MA, USA, 2016; pp. 133–139. [Google Scholar]
- Holden, C.J.; Dennis, T.; Hicks, A.D. Assessing the reliability of the M5-120 on amazon’s mechanical Turk. Comput. Hum. Behav. 2013, 29, 1749–1754. [Google Scholar] [CrossRef]
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).