We have come a long way since the first release of Microsoft Grammar Checker in the 90s, with Artificial Intelligence (AI) massively expanding the possibilities. The so-called language tools and writing assistants now encompass much more than grammar and can help with style with many features such as paraphrasers and summarizers. Some even include features that can generate an entire text, even abstracts, based on input you provide. When thinking about using such tools, factors to consider include not only features and reliability but also the price. The golden rule of shopping, “higher price doesn’t mean better services”, also applies here. While some tools are entirely free, many have limited free functionalities, requiring payment to unlock a full range of features. 

In this post, we provide an overview of current grammar tools, discussing their advantages and disadvantages, as well as considering the potential risks and future developments of these tools. We also propose key questions to consider before using them. For a list of selected language tools that we tested and compared, visit NTK’s guide. 

Everyone has their preferences, but…

Everyone has their own preferences or even demands for these kinds of tools. The aesthetics of design, functionality, reliability, the number of features, and price are among commonly-important factors. For us, one of the important factors is a tool’s interface and clearness, both of which aid in effective and easy orientation and use. This of course relates to the display language; most of the tools we tested were in English, but some of them were in different languages, such as German, French or Spanish (even if they managed to make corrections in English). What could be considered a useful feature for some is the generation of new text such as title or abstract based on an uploaded essay or article. Such text can provide you with inspiration and ideas while drawing your input.

For us, the user experience fell apart with some tools with obvious limitations. When working on a paper or having several full-page documents, copying and pasting everything into very small fragments makes the process inefficient and slower than it could be. Limiting unique features in free versions is one thing, but basic feature limitations, for example, grammar correction, are frustrating. Many grammar correction tools are available with interfaces in languages with many speakers, such as German, French, or Spanish. For languages spoken by fewer people, such as Czech, various tools offer features like synonym generation, punctuation correction, and grammar explanations. However, the Czech grammar correction tools we tested did not meet our expectations.

A quick mention of our favorites

During our testing, some tools particularly captured our attention due to their unique features and performance. Sentence Checkup stood out with its reliable grammar correction and various interesting features. Plus, it’s free, making it a great option for everyone. While evaluating tools, we set up our criteria, and one of them was “first impression”. Some tools had good functionality, but their design or organization was a bit confusing. In this regard, Multilings really shined. Even though it only offers a one-week free trial before requiring payment, it has a plagiarism checker, translation services, and an easy-to-use interface. One more crucial factor for us was price. During our review, we found that some tools can be quite expensive, so it’s up to you to decide whether you want to invest in them or if they’re worth the price. Of course, there are many great tools worth checking out, and you can find a more detailed list sorted by categories here.

What’s the real (not just financial) cost?

Language tools, especially those based on AI, bring to the table not only effectiveness but also fears and risks. What happens with the data we provide? Will our texts be used for AI training? Will anyone have access to a text we uploaded?  Many questions come to mind when dealing with these kinds of AI-based tools. In case you want to be sure, we suggest checking the tool’s data policy, as some tools promise to never store, use, or sell input, while others may use data for different purposes (e.g., AI training). 

Is it AI, or is it us? 

Advanced features can seem to be on the cutting edge of writing ethics. With features such as paraphrasing, one’s input and words fade and the manner of expression changes, even though the meaning stays more or less the same. Where are the limits? Can we use this kind of text for academic writing? Is there a moment when a text is no longer “ours”? These questions have prompted universities and other (often academic) institutions to publish guidelines on generative AI use. Some guidelines allow using generative AI tools in limited circumstances, which they explicitly state; some ban AI completely. Some institutions haven’t even published guidelines yet, which could assist when making decisions about whether to use generative AI or not. The ethical aspects of academic writing (academic integrity) are definitely something we should consider, especially given how fast these technologies are evolving and how quickly tools offer new, creative ways to make our work easier.  

While dealing with these kinds of tools, especially text generation and advanced editing, a question arises when considering modern education methods. Should such tools be used? If so, how? How will they influence the student’s work, and what are the possible consequences? Outcomes and final responsibility now rest with us. 

The downfall? 

Many tools offer their own set of extensions for browsers and MS Word. Still, many just provide online features, which means you have to visit a website and copy/paste the selected document into the tool’s interface. There are exceptions, such as Writefull or Grammarly, which have incorporated features directly into MS Word (which can be useful, mainly because it can improve effectiveness). However, not many of us want to keep returning to a tool’s website and losing time editing text in two separate places. Is there a future for such tools? Could they be more helpful if integrated into text editors or web browsers? Only the future will provide us with better answers.

The question of how many chatbot-generated papers would be handed in by students loomed large over high schools and universities last year. The academic sector too started pondering possible rules and regulations to ensure that current standards would be maintained, especially as regards the quality of research and research assessment, as well as ethical practice more generally. In response to all of this, 2023 brought lively discussions about many aspects of generative AI tools, mostly chatbots. Updated scholarly publication and citation guidelines, as well as changes to student assessment were some of the results. Towards the end of the year, the Faculty of Business Administration at the Prague University of Economics and Business (VŠE) announced the abolishment of Bachelor theses due to the impossibility of ruling out the use of generative AI tools.

Friend or Foe?

This may have been a slightly unexpected step to those who had – by that time in November – started to see generative AI as less of a threat in that department. AI chatbots can be awe-inspiring when it comes to a variety of tasks, but their inability to provide information and sources reliably hardly makes them fit to generate a thesis that would hold up to scrutiny (without profound editing). Large Language Models (LLMs) – in simplified terms, the technology that has made AI chatbots possible – have the significant disadvantage that text is generated based on probability. This means that while grammar is usually correct and style appropriate, factual inaccuracies slip in from time to time. These are referred to as hallucinations. New users, in particular, might not think it necessary to double-check a chatbot’s output, impressed by the overall presentation. (As regards languages with less speakers, such as Czech, hallucinations occur much more frequently. Additionally, grammatical mistakes appear as well, since the LLM has been trained on a smaller dataset.)

           AI chatbots can be especially frustrating when working with references. Those who attempted to use them as search engines would generally be confronted with made-up titles of non-existent papers, including non-functional URLs or DOIs. Even if correct bibliographic information was entered as part of the prompt, the chatbot would usually corrupt it in the text it generated by creatively changing some details. Regarding ChatGPT specifically, its inability to access the internet (in the free version) is an additional hindrance. Of course, ChatGPT is not the only impressive AI chatbot on the market, with Google’s Gemini (previously Bard) currently receiving a lot of attention, in part due to its later release. Gemini, as well as Microsoft Copilot (previously Bing Chat) or Perplexity AI do have access to the internet. The latter two also consistently cite their sources as functional hyperlinks next to their answers (while still lacking the ability to generate hallucination-free text). This makes them resemble Wikipedia, and it is recommended to approach them in a similar way. That means remaining highly wary of the text itself but leveraging the sources they reference.   

           What seems to be the overall consensus among the public, and teachers and instructors more specifically, is that generative AI warrants new approaches to written assignments. However, the situation is far from black-and-white. Surely, some students will give in to the temptation to hand in AI-generated coursework even when told not to do so, which is an issue that needs to be addressed. But when used critically and with guidance, AI chatbots can also play their role in information literacy classes, serving as examples that not everything on the internet should be trusted. Additionally, their potential to save time by laying the groundwork, for instance by producing text that is then expanded by a human (or at least carefully edited), also often comes up in discussions. A brainstorming tool seems to be the most popular designation of the currently available AI chatbots.

Some AI tools are still waiting to be discovered

Perhaps the most important lesson that we can draw from 2023 is that while the world’s leading tech companies have captivated us with their race to release the most successful AI chatbot on the market, there is much more out there worth our attention, especially if we are students, scholars, or librarians. A variety of AI tools designed specifically to aid in academic research have been released – and are continuously being developed – without much notice from the general public, or even students and academics.

Even if we remain in the area of prompt-based searches, meaning that we ask the tool a fully-formed question (as opposed to using keywords, Boolean operators, and filters), there are many exciting options geared to the academic sphere. SciSpace, Elicit, Consensus, and even the above-mentioned Perplexity AI, can all be used to look for scholarly articles. If, on the other hand, you already have a collection of resources in your own digital library, you might be interested in tools that allow you to search for additional, similar papers (and books). You can upload your personal collection into the following apps and decide which of them works the best for you: Research Rabbit, Litmaps, Inciteful, or Connected Papers, among others.

If you would like to learn more, check out our guide to AI tools for academic purposes.

So, have fun exploring in 2024!

“Open notebook” science is a practice in which research scientists record their work online and make it publicly available while conducting research in near-real time. Such research is thus completely open to the public and includes all its aspects such as raw data and any associated material. The approach was described by Bradley (2007), the first to use the term, as “no insider information”. Open notebook science makes the research process transparent and provides unsuccessful, not very significant, or unpublished outcomes (sometimes called “dark data”) to anyone interested (Goetz 2007).

Let’s take a look at the primary advantages of this practice, according to the literature. According to Schapira, Harding and Consortium (2019, p. 3), open lab notebooks can save time, resources, and knowledge. Making the information accessible quickly means that other researchers will be able to build upon the open results, making it possible for others to avoid spending time and resources on redundant experiments (Powell 2016). Open lab notebooks should include detailed protocols to achieve experimental replicability. The necessity of more transparent, replicable experiments has been discussed recently (for example, by Nature 2016 and Wallach et al. 2018). Negative data from unpublished research might additionally provide important insights (Mlinarić et al. 2017; Nimpf and Keays 2020). Open lab notebooks can also give experts a space for discussion, to (for example) spot discrepancies in an experiment and so on.

Early career researchers can use their notebooks to connect with peers and experts in the field. One can also add a link to one’s research notebook in an academic CV when applying to a position in order to showcase technical skills (Schapira, Harding and Consortium 2019).

Possible drawbacks of open notebooks (Harding and Consortium 2019, Sanderson 2008, Schapira 2018, Zirnask 2014):

• Possible data theft (being “scooped”): Risk can be mitigated by using repositories such as Zenodo, which assign a citable DOI (or other citable record) to a notebook.
• Difficulty publishing open notebook results in traditional peer-reviewed journals.
• Influencing other research projects before the research documented in the open notebook is complete and/or well-analysed (this is why there are experimental collaborations without open notebook use with strict rules to avoid data leakage and issues with influencing results).
• “Data deluge”: flooding the information space with a large amount of non-peer-reviewed material.
• Can be difficult without a smooth process: maintaining an open notebook should be executed effectively to avoid wasting too much time (as with a regular lab notebook).

What are the necessary characteristics of open notebooks? Harding (2019, p. 2) notes she designed her notebook “to be discoverable, accessible, clear, and detailed in its presentation, and to permit dialogue between readers and me, and to pave the way for collaborations.”

Examples of platforms for open notebooks:

Picking the right open notebook platform can be daunting. Every scientific field has a different environment with different data collection requirements and different data types (e.g., code, images, equations, value). Several open notebook options are listed below; you may wish to ask your mentor if they have a preference for a particular tool. One could additionally create a blog for an open notebook.

• https://openlabnotebooks.org/

To create an open notebook, contact the coordinator. Data are uploaded and stored on zenodo.org (maintained by CERN as a part of the OpenAIRE initiative). You can also link zenodo files to your ORCID profile.

• https://github.com/

Used by many open sourced software developers, among others. A code repository that allows parallel code editing.

• https://openwetware.org/

Join via a web form. Designated for biology and biological engineering.

• https://jupyter.org/

Free software, open standards, and web services for interactive computing across all programming languages.

Various interfaces (lab book, notebook, hub, Voilá) for sharing outputs.

Use https://nbviewer.org/ to make a jupyter notebook publicly shareable.

• https://open-neuroscience.com/

Open database for neuroscience projects.

If your research, for any reason, cannot be made public, another option for organisation and cooperating/sharing with other researchers or within teams, or managing protocols is the use of electronic laboratory notebooks. There are many options to choose from such as: https://rmarkdown.rstudio.com/, https://www.labarchives.com/, https://www.benchling.com/, and https://www.elabftw.net/ (among others).

Openly published protocols: If you do not wish to share your process but would like to publish and share a protocol which you have designed, a tool like https://www.protocols.io can be considered. After publishing a protocol on protocols.io you obtain a digital object identifier (DOI). A DOI can be used in a manuscript so that it can, if the article is approved for publication in a scientific journal, be published (automatically, if a DOI is used upon article submission) at a later date.

Resources:

Bradley, Jean-Claude. (2007). Open notebook science using blogs and wikis. Nature Precedings. https://doi.org/10.1038/npre.2007.39.1

Goetz, Thomas. (2007). Freeing the dark data of failed scientific experiments. Wired, 15(10). Available from: https://www.wired.com/2007/09/st-essay-3/ 

Harding, Rachel J. (2019) Open notebook science can maximize impact for rare disease projects. PLoS Biol, 17(1). https://doi.org/10.1371/journal.pbio.3000120

Mliarić, Ana, Horvat, Martina. & Šupak Smolčić, Vesna. (2017). Dealing with the positive publication bias: Why you should really publish your negative results. Biochem Med, 27(3). https://doi.org/10.11613/BM.2017.030201

Nature. (2016). Reality check on reproducibility. Nature, 533(7604), 437-437. https://doi.org/10.1038/533437a

Nimpf, Simon & Keays David A. (2020). Why (and how) we should publish negative data. EMBO Rep., 21(1). https://doi.org/10.15252/embr.201949775

Powell, Kendall. (2016). Does it take too long to publish research? Nature, 530, 148-151. https://doi.org/10.1038/530148a

Sanderson, Katherine. (2008). Data on display. Nature, 455(18), 273. https://doi.org/10.1038/455273a

Schapira, Matthieu. (2018). Open lab notebooks to increase impact and accelerate discovery . Springer Nature. Available at: https://researchdata.springernature.com/posts/29655-open-lab-notebooks-to-increase-impact-and-accelerate-discovery

Schapira, Matthieu & Rachel J Harding. (2019). Open laboratory notebooks: good for science, good for society, good for scientists. F1000Res, 8(87). https://doi.org/10.12688/f1000research.17710.2

Wallach, Joshua D, Boyack, Kevin W. & John P. A. Ioannidis. (2018) Reproducible research practices, transparency, and open access data in the biomedical literature, 2015-2017. PLoS Biology, 16. https://doi.org/10.1371/journal.pbio.2006930

Zirnask, Mart. (2014). Are open notebooks the future of science? UT Blog. Available at: https://blog.ut.ee/are-open-notebooks-the-future-of-science/

While such tools are useful, they (in my opinion) do not replace clear thought. I often tell non-native students that if they cannot express themselves well in their native language, none of the English writing tools will help them present their thoughts better in this second language. 

In addition to LanguageTool, my students and colleagues find the following tools of use:

• Grammarly (subscription tool)
• QuillBot (open, particularly helpful in finding alternative phrasing)
• GoogleTranslate
• Glosbe Dictionary (translator for terms in context) 

I tried to find a comparison chart created by universities or libraries for these tools, but was unsuccessful. Various lists of the so-called best tools for 2022 (scroll past the paid content) are available in this sample search.

Stepping backwards, here is a nice subject guide to editing and proofreading in English that includes a nice checklist. 

Curtain University (2021). Editing and proofreading your assignment. https://libguides.library.curtin.edu.au/uniskills/assignment-skills/writing/editing-proofreading 

Read more:

Demeter, E. (2021). Reflecting on Course Redesign: How Faculty Can Measure the Impact of Instructional Changes. Liberal Education Blog. https://www.aacu.org/blog/reflecting-course-redesign-how-faculty-can-measure-impact-instructional-changes

Mudditt, A. (2021). Emerging from the Pandemic: The Future of Work is Now. Scholarly Kitchen. https://scholarlykitchen.sspnet.org/2021/02/09/emerging-from-the-pandemic-the-future-of-work-is-now/

Anderson’s conclusion:

Overall, I find the Cabell’s Blacklist product to be a carefully crafted, honestly managed, and highly useful tool for libraries, faculty committees, and authors.

Read more:

Anderson, R. (2018). Cabell’s Predatory Journal Blacklist: An Updated Review. The Scholarly Kitchen. https://scholarlykitchen.sspnet.org/2019/05/01/cabells-predatory-journal-blacklist-an-updated-review/

There will be a shakedown in the business. Some journals will lose out. Publishing is not a static business — the advent of the preprint server has really changed things, for example. Journals are going to change, and Plan S could have a strong influence.

Else, H. (22 Feburary 2019). Open-access pioneer Randy Schekman on Plan S and disrupting scientific publishing, Nature News Q&A. Available at:  https://www.nature.com/articles/d41586-019-00595-y?utm_source=twt_nnc&utm_medium=social&utm_campaign=naturenews&sf208184622=1