Periodic Reporting for period 1 - PLanTra (Plain language for financial content: Assessing the impact of training on students' revisions and readers' comprehension)
Reporting period: 2020-08-01 to 2022-07-31
The PLanTra project addressed the need for clear financial and business communication. With more and more people reading specialised content in order to make decisions that have a long-lasting impact on their financial wellbeing, the importance of plain language cannot be overestimated. Training current and future writers of financial and business documents in the use of plain language is therefore key to ensure that communications are delivered clearly and effectively to the lay audience.
This fellowship focused on content about corporate social responsibility, namely the extent to which companies invest in the working conditions of their employees, the environment, and society. As more and more customers prefer to buy products from socially responsible companies, information about corporate social responsibility can influence purchase decisions. However, there is empirical evidence that such information is rather difficult to read.
Against this background, we developed plain language training specifically tailored to corporate content on social responsibility, and we tested its impact on: (i) the process by which business students simplify texts; and (ii) the characteristics of their texts. We answered the following research questions:
• Does plain language training have an impact on how business students simplify financial texts?
• Does plain language training have an impact on the resulting comprehensibility of texts?
Our results have led to the following conclusions:
• The writing processes of university students show similarities with those of more experienced writers;
• The majority of second-language university students prefer to rewrite texts from scratch when asked to simplify;
• In the text simplification process, it is possible to identify two macro-level phases — a first phase during which students produce most of the content, and a second phase characterised by reading/editing behaviour;
• The duration, frequency, and location of writing pauses is linked with readability.
• Our plain language training did not influence the way in which second-language university students paused, revised, or looked up sources. However, it helped students with lexical choices and it provided them with stepwise guidance on how to approach the text simplification task.
• Our plain language training led students to produce texts that were shorter and syntactically simpler. However, we found no impact on high-level text characteristics (such as cohesion). This might explain why comprehensibility as assessed by lay readers was not higher.
This fellowship focused on content about corporate social responsibility, namely the extent to which companies invest in the working conditions of their employees, the environment, and society. As more and more customers prefer to buy products from socially responsible companies, information about corporate social responsibility can influence purchase decisions. However, there is empirical evidence that such information is rather difficult to read.
Against this background, we developed plain language training specifically tailored to corporate content on social responsibility, and we tested its impact on: (i) the process by which business students simplify texts; and (ii) the characteristics of their texts. We answered the following research questions:
• Does plain language training have an impact on how business students simplify financial texts?
• Does plain language training have an impact on the resulting comprehensibility of texts?
Our results have led to the following conclusions:
• The writing processes of university students show similarities with those of more experienced writers;
• The majority of second-language university students prefer to rewrite texts from scratch when asked to simplify;
• In the text simplification process, it is possible to identify two macro-level phases — a first phase during which students produce most of the content, and a second phase characterised by reading/editing behaviour;
• The duration, frequency, and location of writing pauses is linked with readability.
• Our plain language training did not influence the way in which second-language university students paused, revised, or looked up sources. However, it helped students with lexical choices and it provided them with stepwise guidance on how to approach the text simplification task.
• Our plain language training led students to produce texts that were shorter and syntactically simpler. However, we found no impact on high-level text characteristics (such as cohesion). This might explain why comprehensibility as assessed by lay readers was not higher.
The first year of the project was centered around Work Package 1 (development of training) and Work Package 2 (experiment on simplification process).
As part of Work Package 1, the work involved:
• Selection of business content that would be the focus of the training: we opted for corporate social responsibility;
• Selection of the hosting environment for the training: we decided to use the online writing center Calliope;
• Development of the training: for the experimental group (who received the module on plain language applied to corporate social responsibility), we opted for a selection of plain language principles. For the control group, we focused on the definition, the evolution, and the communication of corporate social responsibility;
• Pilot testing and refinement of training.
As part of Work Package 2 (experiment on simplification process), the work in the first year involved a small-scale data collection aimed at testing the experimental set-up. We recruited three students and one plain language expert, we asked them to participate in our plain language training and to simplify a document about corporate social responsibility. We collected data using keystroke logging and screen recording.
The second year of the project was centered around Work Package 2 (experiment on simplification process) and Work Package 3 (experiment on simplification product).
The activities carried out as part of Work Package 2 in the second year involved large-scale data collection. We collected keystroke logging data from 47 Master students. For a sub-set of these participants, we also collected screen recordings and retrospective interviews. We randomly divided the students into an experimental and a control group. The experimental group would receive our plain language training, while the control group would receive our module on the topic of corporate social responsibility. The collected data allowed us to understand students' pausing behaviour, their use of sources, and their simplification strategies. The training materials and the texts were in English, while the students were native speakers of other languages. Therefore, we applied a second-language dimension to the analysis of our data.
The activities carried out in Work Package 3 involved the analysis of the comprehensibility of the texts produced by the students. 51 readers evaluated the simplified texts and commented on their comprehensibility. We also measured comprehensibility automatically with the computational tool Coh-Metrix.
The outputs from this project were shared through the following activities (Work Package 4):
• Dissemination:
o publication of the article “Accessible communication of corporate social responsibility: Development and preliminary evaluation of an online module” in Business and Professional Communication Quarterly;
o submission of the three articles currently under review;
o an article currently in preparation;
o six conference presentations.
• Communication:
o PLanTra project website;
o Social media accounts (Twitter and LinkedIn);
o PLanTra newsletter;
o Participation in the Dag van de Wetenschap (Science Day).
• Exploitation:
o The PLanTra training materials are freely available online for re-use and adaptation;
o The PLanTra dataset is available on Zenodo for follow-up and replication studies.
As part of Work Package 1, the work involved:
• Selection of business content that would be the focus of the training: we opted for corporate social responsibility;
• Selection of the hosting environment for the training: we decided to use the online writing center Calliope;
• Development of the training: for the experimental group (who received the module on plain language applied to corporate social responsibility), we opted for a selection of plain language principles. For the control group, we focused on the definition, the evolution, and the communication of corporate social responsibility;
• Pilot testing and refinement of training.
As part of Work Package 2 (experiment on simplification process), the work in the first year involved a small-scale data collection aimed at testing the experimental set-up. We recruited three students and one plain language expert, we asked them to participate in our plain language training and to simplify a document about corporate social responsibility. We collected data using keystroke logging and screen recording.
The second year of the project was centered around Work Package 2 (experiment on simplification process) and Work Package 3 (experiment on simplification product).
The activities carried out as part of Work Package 2 in the second year involved large-scale data collection. We collected keystroke logging data from 47 Master students. For a sub-set of these participants, we also collected screen recordings and retrospective interviews. We randomly divided the students into an experimental and a control group. The experimental group would receive our plain language training, while the control group would receive our module on the topic of corporate social responsibility. The collected data allowed us to understand students' pausing behaviour, their use of sources, and their simplification strategies. The training materials and the texts were in English, while the students were native speakers of other languages. Therefore, we applied a second-language dimension to the analysis of our data.
The activities carried out in Work Package 3 involved the analysis of the comprehensibility of the texts produced by the students. 51 readers evaluated the simplified texts and commented on their comprehensibility. We also measured comprehensibility automatically with the computational tool Coh-Metrix.
The outputs from this project were shared through the following activities (Work Package 4):
• Dissemination:
o publication of the article “Accessible communication of corporate social responsibility: Development and preliminary evaluation of an online module” in Business and Professional Communication Quarterly;
o submission of the three articles currently under review;
o an article currently in preparation;
o six conference presentations.
• Communication:
o PLanTra project website;
o Social media accounts (Twitter and LinkedIn);
o PLanTra newsletter;
o Participation in the Dag van de Wetenschap (Science Day).
• Exploitation:
o The PLanTra training materials are freely available online for re-use and adaptation;
o The PLanTra dataset is available on Zenodo for follow-up and replication studies.
The PLanTra project has been the first to investigate text simplification in a second language by combining a process and a product perspective. Furthermore, no other project has developed and tested the impact of a plain language module tailored to content on corporate social responsibility. As such, this fellowship has pedagogical implications. We have shown that providing second-language students with training on plain language does not seem to alter their overall writing process, but might assist them in becoming more aware of which textual aspects need consideration while simplifying a text. Also, we have observed that plain language training leads second-language writers to address issues mainly at the sentence level, while cohesion does not seem to benefit from the training. Teachers might keep this aspect in mind and develop instruction that specifically addresses high-level text characteristics. Additionally, we have shown that more frequent and more prolonged pausing can be beneficial for text readability. Therefore, it might be worth fostering students’ understanding of the impact that pausing can have on their texts. These conclusions are important when considering that the ability to produce clear communications is increasingly sought out in the workplace, especially in a corporate environment, where transparency about social responsibility can have a beneficial effect on profits. Therefore, receiving effective plain language training can increase the employability of students.