#AcWriMo: Overcoming Academic Writing Blocks

In the second of our academic writing-themed posts during #AcWriMo this year, Dr Lizzie O’Connor, Postgraduate Community Engagement Manager in the University Graduate School, acknowledges that we all struggle to write at times, and suggests some strategies to overcome this.

Writing is one of the most important parts of our academic lives, but it can also be one of the most fraught. Dorothy Parker’s words that joy comes not in writing, but in having written, can ring very true as we stare at a blank screen or an unedited paper for hours on end, waiting for the motivation to start, continue, or finish our writing.

Photo by Henry & Co. on Pexels.com

Many of us have an ‘ideal’ writing practice in our mind, such as daily writing targets, a fluent style, or time spent writing, and many of us see this ideal defeated by procrastination, lack of confidence, distraction. It can leave us feeling disappointed, frustrated, and even ashamed, which feeds into a cycle of worrying about our writing and – crucially – not fully engaging with our writing practice. What techniques can we employ to overcome these blocks, and build good writing habits? What stops us from writing in the way we’d really like to?

Silencing Our Inner Critic

We all have an inner critic, and for many of us this critic comes out most fiercely when we are trying to write. The inner critic will tell you that your writing is terrible, that none of it is good enough in quality or quantity, and becomes the main culprit for our procrastination. Even great writers struggle with this: when Marlon James won the Booker prize, he attributed his success to getting up earlier than his inner critic.

One of the most powerful techniques to manage the inner critic is to confront and recognise this sabotaging voice, rather than following the impulse to ignore it or stop writing altogether. I like the following suggestion from best-selling author Cathy Rentzenbrink to talk to your inner critic in order to disarm it:

Being curious about it or getting to know it can help. […] Put ‘What do you want?’ at the top of the page and just let your hand move. If you find out the motivation you can offer some reassurance. If we know its intention, we can say, ‘Thank you. My today self no longer needs you to protect me.’ Or you can say, ‘I know you are trying to help, but when you hit me over the head with a stick and tell me I’m rubbish, it paralyses me and then I can’t get stuff done, so could you be a bit nicer?’ Or you can imagine yourself locating the volume switch in your head and turning it down.

Target-setting and Procrastination

We often accuse ourselves of laziness when we procrastinate, but like the inner-critic, procrastination is more often an anxiety-based response: your brain shying away from a task that feels frightening, is overwhelming, or that has high stakes.

You’ve probably heard of target-setting as a technique to overcome procrastination. It’s something the Research Skills Team use in their in-person Shut Up and Work sessions, and that we use in the University Graduate School’s online version.

A method I used in writing my own PhD thesis was setting anti-targets: writing targets so low in effort and ambition they scraped the bare minimum of what I could write each day, such as, say, writing 300 words. It worked because it was so unintimidating: I could write freely, and in my busiest days could always squeeze it in, keeping up a momentum of daily writing. There was no voice in my head berating me for taking a lunch break, or daydreaming, or worrying. It set a positive pattern of self-worth: instead of starting the next day feeling like a failure because I didn’t meet my targets or wasn’t productive enough the day before, I could start it positively, feeling like a success. Feeling as though I could write and could accomplish, I did. And those 300 words added up more than you would think!

Ultimately, the only way to build a writing practice is to write. It’s a vital part of our work as academics, but also a lifelong skill in expression and the process of drafting. The tips above rely wholly on self-compassion: take the pressure off, be kind to yourself, get words (any words! Terrible words!) on the page, and join Dorothy Parker in the joy of having written.

#AcWriMo: Critical Engagement with the Literature

November is #AcWriMo! In the first of our academic writing-themed posts this month, Dr Kate Spencer-Bennett, Academic Skills Advisor in the Academic Skills Centre, considers how we might approach the literature in our field in a critical way.

You’ll often hear it said that good academic writing involves a ‘critical engagement’ with the literature. And you probably know that an effective literature review involves something more than a summary of everything that has been said on the topic. So, if Academic Writing Month (or #AcWriMo) is inspiring you to settle down to your desk, then you might be asking how you can comment on your reading in critical ways.

There are, of course, many ways to talk about your reading, but I often think that comments fall into one of three broad groups – the evaluative comment, the analytical comment, and the connection-making comment. Let’s consider each.

Evaluative comment

This type of comment reveals a critical engagement with the literature because it assesses the value of a piece of research. It asks, what are the strengths and weaknesses of the work? What is interesting, useful, or valuable about this article, book, or report? Where are its limitations? Thinking and reading in these ways might lead you to a more critical approach with your writing.

This kind of comment might begin, ‘This work is valuable because …’

Analytical comment

This type of comment seeks to look carefully at something and offer an interpretation. You might pick out one element of a text for close inspection. Is there an interesting argument being made? Is there a particular point, statistic, or telling phrase which you can point to? What does it suggest about the writer’s position?

This kind of comment might begin, ‘The use of the phrase x suggests …’

Connection-making comment

This type of comment aims to draw connections between different texts. Rather than summarising the research piece by piece, connection-making comments find agreements and disagreements in the literature. When you are reading, you could ask yourself how a particular journal article, for example, responds to what has been said on the topic before. What has been said since?

Or, alternatively, you might look at broader patterns within the research in your field. Can you begin to group what you have read by theme? Where do you see harmony in approach or viewpoint? Where are the tensions? These kinds of questions can lead you to synthesis in your writing. You bring different elements together to make something new.

This kind of comment might begin, ‘This work aligns with …’

When evaluating, analysing, and making connections you demonstrate a close reading of the literature and, in subtle ways, reveal your own perspective. Whether you feel that you are sometimes too descriptive, or just want to make sure that your own voice comes through as you discuss the existing research, perhaps these ideas could get you started next time you sit down to write.

AI: what are the risks?

Continuing our mini-theme on artificial intelligence, Alex Fenlon, Head of Copyright and Licensing in Library Services, addresses some of the concerns around these new and emerging tools in this in depth post. He highlights some of the issues that users need to be aware of when trialling these new tools in research.

New AI tools, including those mentioned in our previous post on this topic, are an exciting new development in research and present us with some fantastic opportunities. However, there are a few caveats which researchers need to be aware of before AI tools are included as a central part of a researchers’ toolkit.

Photo by Redrec u00a9ufe0f on Pexels.com

Reproducibility

Reproducibility, replicability, stability, validity, and consistency of results are a key part of research across all disciplines. Researchers must be able to repeat their research to ensure their results are valid and stand up to the scrutiny and rigour that peer review requires. Anything that leads to an inability to repeat a method or replicate results will undermine any findings and outcomes.

The use of some AI tools in research today raises precisely these concerns: whether the consistency of results will be maintained over time and whether identical inputs will lead to identical outputs. If you ask a question today, will you get the same result tomorrow, next week, in a couple of years time, or when you’re thinking about publishing your research?

AI tools are developing at a rapid pace, with enhancements to functionality and quality being released at breath-taking speed. Researchers will need to ensure that the precise version of a tool used is communicated and potentially remains available for interrogation as research progresses. Along with these enhancements, some tools learn and evolve over time and this too will potentially impact on the reproducibility of research. This shifting and uncertainty in responses could have significant impact, and it’s key that researchers do not forget their literacy skills and become overly reliant on these tools.

Exploring AI Tools for Researchers

Following on from our last post on Potential Uses of ChatGPT in Research, this in depth post by James Barnett, Research Skills Advisor from the Research Skills Team in Library Services, covers wider emerging AI tools for research.

What did you make of our previous post, written by ChatGPT in response to a couple of prompts? Did you think the response was thorough and authentic? Did you feel that points being made about the potential uses for ChatGPT were backed up by enough detail for you to be convinced?

How do you feel about content that is in any way intended as being advisory to the research community being generated by an AI tool – uneasy, or just a logical technological progression?

An image of a fantasy world generated by DeepAI.

Whatever your answers to these questions, there can be no doubt that the release of ChatGPT in November 2022 has heralded a sharp focus on the role that generative Artificial Intelligence (AI) tools – and developments in AI more broadly – will have in the way we live our personal and professional lives going forward. In some ways, it’s easy for us all to get wrapped up in the narrative that these developments should be seen as at best mildly concerning and at worst downright frightening – particularly when news breaks that leading scientists and technologists are signatories to an open letter calling for the further development in advanced AI tools to be paused, or that the so-called “godfather of AI”, Geoffrey Hinton, has resigned from Google in order to warn of the dangers of further developments in the field.

While wrestling with profound questions around the role and risks of recent developments in AI are undoubtedly necessary as we adapt to them at a societal level (and the case has already been made for the significant role the academic research community should play in this), we should also balance this up with taking account of the wonderful opportunities these developments might bestow – in the context of academic research generally, and the workflows and processes associated with the research cycle in particular. For example, one of the frustrations about the scholarly communications landscape that we frequently hear in the Research Skills Team is that everything moves too slowly – that the time gap between generating findings and getting them published can take years. There is the great potential to use AI to cut these inefficiencies. Using AI to speed up peer review is one area that has already been identified from the publisher perspective, but there are many ways that AI tools can be useful to researchers to augment their existing processes in a way that saves the researcher time.

Potential Uses of ChatGPT in Research

This post (except for this editor’s note) was written entirely by ChatGPT, in response to a couple of prompts about the potential and appropriate use of ChatGPT in research. The text is presented unedited (except for removing sub-headings and combining responses to two separate prompts). Note that it proved to be non-trivial to get ChatGPT to produce suitable text – there were a number of network errors which interrupted some responses, and ChatGPT wasn’t great at working to suggested word counts.

As a postgraduate researcher, you’re always looking for ways to streamline your work, increase your productivity, and produce high-quality research outcomes. One tool that may help you achieve these goals is ChatGPT, a language model based on GPT-3.5 architecture.

The ChatGPT logo

ChatGPT is a powerful tool that can assist you in a variety of research tasks, from generating ideas and hypotheses to performing complex analyses and interpreting data. In this blog post, we’ll explore some of the potential uses of ChatGPT in research and explain how it can benefit postgraduate researchers.

One of the most challenging aspects of research is coming up with new ideas and hypotheses. With ChatGPT, you can generate a wide range of ideas and hypotheses based on the topic of your research. By inputting a brief summary of your research area or question, ChatGPT can generate several ideas that you can explore further. You can also use ChatGPT to refine your research question or problem statement, making it more specific and focused.

Writing a research proposal can be a daunting task, especially if you’re not sure where to start. ChatGPT can help you write a compelling and well-structured research proposal by providing you with relevant information, suggestions, and tips. You can input your research question, methodology, and desired outcomes, and ChatGPT will generate a proposal outline that you can use as a starting point.

Once you’ve collected data, you need to analyze it to draw meaningful conclusions. ChatGPT can help you analyze your data by suggesting statistical tests, visualizations, and other methods to explore your data. By inputting your data and research question, ChatGPT can help you identify patterns, trends, and relationships in your data, allowing you to make evidence-based conclusions.

Interpreting research results can be challenging, especially if you’re dealing with complex data or statistics. ChatGPT can help you interpret your results by providing you with explanations, definitions, and examples of statistical concepts and methods. By inputting your results and research question, ChatGPT can help you understand the implications of your research findings and draw meaningful conclusions.

Finally, ChatGPT can assist you in writing research papers by providing you with suggestions for sentence structure, grammar, and syntax. You can input your research question, methodology, and findings, and ChatGPT will generate a draft paper that you can revise and edit. This can help you save time and effort in writing a high-quality research paper.

In conclusion, the use of ChatGPT can greatly benefit postgraduate researchers in various stages of their research projects, from generating ideas to writing research papers. However, it is important to acknowledge the ethical implications and appropriate use of this technology. Researchers must ensure that they are not plagiarizing or relying too heavily on ChatGPT-generated content. Additionally, researchers must take responsibility for ensuring that the data they input into ChatGPT is accurate and ethical, and that the outputs generated align with the values of scientific research. As with any tool, it is important to use ChatGPT in a responsible and ethical manner, and to ensure that it supports, rather than replaces, the skills and expertise of the researcher.