The Future of Coding: A Comparison of Hand-Coding and Three Types of Computer-Assisted Text Analysis Methods

Nelson, Laura K.; Burk, Derek; Knudsen, Marcel; McCall, Leslie

doi:10.1177/0049124118769114

Cited by 129 publications

(107 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The coding methods used were hand coding, machine learning, or a combination of the two. Compared with hand coding, machine learning can rapidly code large amounts of data; however, hand coding undertaken by humans may more accurately discriminate the complexities and subtleties of language [54]. Although hand coding can be subject to individual bias, the development of codes grounded in literature and achieving acceptable levels of inter-rater reliability can assist to reduce this [55].…”

Section: Discussionmentioning

confidence: 99%

The Messages Presented in Electronic Cigarette–Related Social Media Promotions and Discussion: Scoping Review

et al. 2019

View full text Add to dashboard Cite

Background There has been a rapid rise in the popularity of electronic cigarettes (e-cigarettes) over the last decade, with growth predicted to continue. The uptake of these devices has escalated despite inconclusive evidence of their efficacy as a smoking cessation device and unknown long-term health consequences. As smoking rates continue to drop or plateau in many well-developed countries, transnational tobacco companies have transitioned into the vaping industry and are now using social media to promote their products. Evidence indicates e-cigarettes are being marketed on social media as a harm reduction alternative, with retailers and manufacturers utilizing marketing techniques historically used by the tobacco industry. Objective This study aimed to identify and describe the messages presented in e-cigarette–related social media (Twitter, YouTube, Instagram, and Pinterest) promotions and discussions and identify future directions for research, surveillance, and regulation. Methods Data sources included MEDLINE, Scopus, ProQuest, Informit, the Journal of Medical Internet Research , and Google Scholar. Included studies were published in English between 2007 and 2017, analyzed content captured from e-cigarette–related social media promotions or discussions, and reported results for e-cigarettes separately from other forms of tobacco and nicotine delivery. Database search ceased in October 2017. Initial searches identified 536 studies. Two reviewers screened studies by title and abstract. One reviewer examined 71 full-text articles to determine eligibility and identified 25 studies for inclusion. This process was undertaken with the assistance of the Web-based screening and data extraction tool—Covidence. The review was registered with the Joanna Briggs Institute (JBI) Systematic Reviews database and followed the methodology for JBI Scoping Reviews. Results Several key messages are being used to promote e-cigarettes including as a safer alternative to cigarettes, efficacy as a smoking cessation aid, and for use where smoking is prohibited. Other major marketing efforts aimed at capturing a larger market involve promotion of innovative flavoring and highlighting the public performance of vaping. Discussion and promotion of these devices appear to be predominantly occurring among the general public and those with vested interests such as retailers and manufacturers. There is a noticeable silence from the public health and government sector in these discussions on social media. Conclusions The social media landscape is dominated by pro-vaping messages disseminated by the vaping industry and vaping proponents. The uncertainty surrounding e-cigarette regulation expressed within the public health field appears not to be reflected in ongoing social media dialogues and highlights the need for public health professionals to interact with the public to actively in...

show abstract

Section: Discussionmentioning

confidence: 99%

The Messages Presented in Electronic Cigarette–Related Social Media Promotions and Discussion: Scoping Review

et al. 2019

View full text Add to dashboard Cite

show abstract

“…Like topic modeling, content analysis shows an extremely meaningful trend with a positive slope. For more details about content analysis, refer to [139]- [144].…”

Section: The Second Category Represented Common Research Paperrelatedmentioning

confidence: 99%

Twitter and Research: A Systematic Literature Review Through Text Mining

et al. 2020

View full text Add to dashboard Cite

Researchers have collected Twitter data to study a wide range of topics. This growing body of literature, however, has not yet been reviewed systematically to synthesize Twitter-related papers. The existing literature review papers have been limited by constraints of traditional methods to manually select and analyze samples of topically related papers. The goals of this retrospective study are to identify dominant topics of Twitter-based research, summarize the temporal trend of topics, and interpret the evolution of topics withing the last ten years. This study systematically mines a large number of Twitter-based studies to characterize the relevant literature by an efficient and effective approach. This study collected relevant papers from three databases and applied text mining and trend analysis to detect semantic patterns and explore the yearly development of research themes across a decade. We found 38 topics in more than 18,000 manuscripts published between 2006 and 2019. By quantifying temporal trends, this study found that while 23.7% of topics did not show a significant trend (P => 0.05), 21% of topics had increasing trends and 55.3% of topics had decreasing trends that these hot and cold topics represent three categories: application, methodology, and technology. The contributions of this paper can be utilized in the growing field of Twitter-based research and are beneficial to researchers, educators, and publishers. INDEX TERMS Literature review, social media, survey, text mining, topic modeling, Twitter.

show abstract

“…The first, most general dimension for evaluating any algorithm is determining the correctness of its results. There are many different measures for evaluating labeling or classification algorithms (Nelson et al 2018). In general, commercial labeling systems present users with only predicted positive labels (e.g., "there are cats in this photograph") and not predicted negative labels (e.g., "there are no children").…”

Section: Evaluating Gcvmentioning

confidence: 99%

“…The overall accuracy of an algorithm such as GCV is not the only important measure, however. As Nelson et al (2018) showed, sometimes the measures of correctness for individual categories and labels are more important for sociological analysis and can lead to further insights about the data. We test this with gender.…”

Section: Bias In Identificationmentioning

confidence: 99%

Diagnosing Gender Bias in Image Recognition Systems

Schwemmer

Knight

Bello-Pardo

et al. 2020

Socius: Sociological Research for a Dynamic World

View full text Add to dashboard Cite

Image recognition systems offer the promise to learn from images at scale without requiring expert knowledge. However, past research suggests that machine learning systems often produce biased output. In this article, we evaluate potential gender biases of commercial image recognition platforms using photographs of U.S. members of Congress and a large number of Twitter images posted by these politicians. Our crowdsourced validation shows that commercial image recognition systems can produce labels that are correct and biased at the same time as they selectively report a subset of many possible true labels. We find that images of women received three times more annotations related to physical appearance. Moreover, women in images are recognized at substantially lower rates in comparison with men. We discuss how encoded biases such as these affect the visibility of women, reinforce harmful gender stereotypes, and limit the validity of the insights that can be gathered from such data.

show abstract

The Future of Coding: A Comparison of Hand-Coding and Three Types of Computer-Assisted Text Analysis Methods

Cited by 129 publications

References 51 publications

The Messages Presented in Electronic Cigarette–Related Social Media Promotions and Discussion: Scoping Review

The Messages Presented in Electronic Cigarette–Related Social Media Promotions and Discussion: Scoping Review

Twitter and Research: A Systematic Literature Review Through Text Mining

Diagnosing Gender Bias in Image Recognition Systems

Contact Info

Product

Resources

About