I believe that designing academic posters is similar to academic writing. You may start with some basic template, stick to the script, and prepare something presentable, something that “fits in”. But, over time, you start to find your own approach and unique style, the goal now being to stand out, rather than to fit in. For me, I feel like I’m learning or trying something new every time I’m designing a new poster, something that is very evident in the timeline of my own previous posters.

The key quality of a good academic poster is that it gives its audience a reason to stop and look at it. Like I said, poster sessions are truly awful experiences. Although there’s a ton of people around with seemingly nothing really to do, only few people actually bother to look at you or your poster. Designing a poster that can draw people in, and that will give you opportunities to talk about your work, can make this experience less stressful for you. A poster really only should be a conversation starter, a networking tool. The actual knowledge dissemination happens in talks with your audience.

In the past, I have used little tokens that thematically fit my research to pass out to people walking by. For my poster on sleep, screen time and executive functions, for example, I handed out tea bags, with a QR code and link to my study. At a conference last year I asked people to write on my poster. Good graphics or plots, however, can also draw people in and, what’s more, they give you a chance to talk about them.

My Process

Structure

The way I like to design academic posters, is to start with some kind of structure. In the past, that meant sketching out boxes on a piece of paper, arranging and re-arranging them. For my two most recent papers, however, I simply googled academic posters and looked for structures and designs I liked. Here’s a recent examples for a poster I designed for our SCOOT Study. You can clearly tell how the structure inspired the final design.

Font

Next, I like to look for a color scheme and font. I’m really proud of the fonts I have amassed on my computer. From And This Happend to Ice Cream Grande to Zoika Font. For a poster, what I’m looking for is a clean, sans-serif font, which has multiple weight levels in between Regular and Bold. This gives me a lot of flexibility in designing and highlighting text. I generally like Avenir Next, but I have been using D-Din for my most recent posters.

Colour Palette

Selecting a colour scheme is equally important and also a fun process. Some research projects already have a set colour palette (e.g., the ECHO project). For others, we can get more creative. I like pastell colours, however, for posters I tend to go for bold, strong colours with clear contrasts. I like this website, which lets you explore colour palettes and generate your own.

Designing

I tend to use Affinity Photo (Vers. 1.10.8) for any of my photo editing and designing processes. Canva, of course, is also a great program, particularly as the full version, however, it’s nice to have a little more control over the editing process, which Canva cannot really give you. PowerPoint is also easy to use, however, it does not give you any actual editing capabilities, so the end product is, most of the time, just basic.

It’s important to set the correct file size right at the start to avoid having to resize anything later on. Most conferences I have ever been to always require A0, which is 841x1189mm (or 33.11x46.81 in). That’s about the only thing that’s fixed at the beginning of the design. Everything else, in my process at least, is purely trial and error. For my most recent design, for example, I spent two hours making a chart, only to realise later on that it was completely unnecessary.

Having the overall structure in mind, I sketch out specific sections on paper. At this point, I don’t have any text written yet. Instead, I simply use scribbly lines to first plot out the entire poster. Usually, about halfway through the designing process, I realise that there’s not enough space on the poster, so I always end up cutting a lot. As you can see, not all of my sketches make it onto the final poster, but they clearly influence my designing process.

Text

The one comment you can trust you’ll always get when it comes to designing posters is that there’s too much text. It is an impossible task to put your full 12,000 word journal article into one simple A0 poster - so you shouldn’t even try. A poster really only is a conversation starter - the actual knowledge dissemination happens in conversation between you and your audience. Text that may take away attention from your conversation, therefore, is only a distraction.

I like to think of the text on my poster as speaking notes for me when engaging with my audience. They are the main points I want to hit when talking about this piece of research. Therefore, I only try to use bullet points and keep my arguments as broad as possible. Yes, sometimes larger chunks of text are unavoidable, specifically for qualitative research. Still, try to keep it functional and minimal.

Visuals and Plots

What’s more important than the text, are the visuals on your poster. Just like the text, they are not supposed to stand on their own, but instead, allow you to use them in conversation with your audience to drive your main arguments home. Basically, they are supposed to give you something to point to. I use R in all my analysis and are, therefore, also creating all of my plots using ggplot2 or base R commands. When putting my plots on my poster, however, I tend to re-adjust or re-create them inside my editing program to give me a bit more control of axes labels and positioning. In my recent poster, for instance, I completely changed up the forest plot from our recent meta-analysis by replacing study names with numbers, changing the colour scheme to fit my poster, rotating it vertically, and making the bars slightly larger.

I’m definitely not of the opinion that everything on your poster needs to be functional. I always used little graphics, or visually pleasing effects as a way to make my posters a bit more attractive. 3D effects, or little shadows, for example, are great. Also, slightly rotating text and putting a box around it, makes it stand out and important. A clip-art person pointing towards a piece of text can also be a nifty trick. If anything, a poster should be fun - fun for you to design and talk about, and fun for your audience to look at.

Useful tools and resources

• Coolors: As noted, I like working off colour palettes and Coolors is a great website to get some inspiration from. I tend to use the “Explore Palettes” function, rather than to generate my own. I find it’s also an interesting look into my mental state at the time to see which palette I’m most drawn to at the time.
• Noun Project: I use icons and graphics from the Noun Project in most of my presentations and posters. The way I usually work is that I first look for a designer that I like, and then sticking to their designs within one poster or presentation to ensure all icons and graphics are in the same style.
• QR Code & TinyURL: QR codes are like newsletter subscriptions. They are very easy to get, but terribly difficult to keep track of. There’s nothing worse than realising that the printed QR code on your poster has expired. I like to use TinyUrl, which is a nifty website to shorten URLs and create QR codes. As long as you keep your account, your links and codes also remain active
• GoogleDocs: I always link the QR code on my poster to a GoogleDoc, where I can compile a list of resources, references, as well as link my e-mail address. What’s more, I can change this GoogleDoc anytime, without having to change the QR code. This is quite handy and just one more way to personalise an academic poster.

That’s all.

What’s your process of designing posters? Where do you start, what are your main challenges, what tools do you use? Also, if you have a poster that you are really proud of, please share it with me. I’m hoping to put together a library of fun posters for future inspiration and learning.

Is this useful to you? What else should be added? Let me know!

Yes, of course, transcribing, the hard way, is a great way to really get to know your data, but at the same time, it also kind of makes you hate your data. And your voice. It’s no fun. This is why I love Zoom transcripts. It minimizes the time I have to listen to my own voice. And yes, it does have its little hiccups, like not fully appreciating my German accent at times, but overall, it speeds up the transcription process immensely.

Still, there’s all this junk in the transcript Zoom produces that requires further cleaning and formatting before the transcript is actually ready to be shared with participants (e.g., to check, add to, or approve), and to be analyzed. You can do this manually within your TextEdit, Word or even Excel. But maybe, this time may be better spend actually getting to know your data rather than chasing empty lines and unnecessary clutter?

What’s all this junk in my Zoom transcript?

Zoom’s transcripts are meant to be used alongside their recorded video. That is why Zoom does not produce a simple text file, but instead saves the transcript as a VTT, or Video Text Track, file. This type of file is commonly used for subtitling of videos, which is basically what we do when we ask Zoom to produce a transcript. Consequently, the transcript does not only contain the text of all your thoughtful questions and your participants’ insightful remarks, but also their corresponding number and time-stamps, which neatly map onto the video-recording from your interview.

While this may be a nice feature in case you ever plan to host a interview watch-party, it may not really be helpful to you in your qualitative research pursuits, as the transcript (1) breaks up speech into separate lines, and (2) contains all this junk, namely time-stamps and numbering, that are not relevant to you. We want to know all the smart things our participants say. We don’t really care at which hundreds of a second they say it. The following R code, therefore, removes all this junk, so that we can focus on what actually matters.

How do I clean my Zoom transcript using R

VTT files are basically just text files, which can be read into R as a character vector either using Base R commands (i.e., readLines()), or using the readr package within the tidyverse. So, just like we would with any quantitative dataset, we can load our qualitative data, i.e., the transcript downloaded from Zoom, into R.

#load library
library(readr)

#read in transcript
transcript <- read_lines(file.choose(),skip = 1, skip_empty_rows = TRUE)

Within this piece of code, we are already getting rid of some of the junk Zoom inserts into our transcripts, namely empty rows (i.e, skip_empty_rows), as well as the first line in the VTT file (i.e. skip = 1), which always starts with “WEBVTT”. The junk remaining now are just the time-stamps, and the line numbers.

transcript <- transcript[!grepl("^\\d+$", transcript) &  #remove lines that contain only numbers (e.g, 330)
                         !grepl("-->", transcript)]      #remove timestamp lines (e.g., 00:31:56.000 --> 00:31:58.360")

This code uses the grepl() function within R to get rid of (1) lines that contain only numbers, and (2) lines that contain time-stamps. The word grepl means “grep logical” and its function allows R to search for distinct patterns of characters within strings based on regular expression syntax (e.g., \d for numbers) or specific text input we provide (e.g., "-->")

These two lines of code already remove all of the junk Zoom inserts into our valuable qualitative data. However, Zoom also breaks interview responses apart, to allow for subtitling line by line. This is unacceptable for us, as we would like to read, and analyze our participants’ responses in full, rather than separate chunks. So, we’ll need to take this one step further and merge consecutive lines from the same speaker.

#merge consecutive lines
cleaned_transcript <- c()
current_line <- transcript[1]

for (i in 2:length(transcript)) {
  
  #extract speaker label (everything before colon)
  prev_speaker <- sub(":.*", "", current_line)
  this_speaker <- sub(":.*", "", transcript[i])
  
  if (prev_speaker == this_speaker) {
    #remove speaker label from new line and append text
    text_only <- sub("^[^:]+:\\s*", "", transcript[i])
    current_line <- paste(current_line, text_only)
  }
  else {
    cleaned_transcript <- c(cleaned_transcript, current_line)
    current_line <- transcript[i]
  }
}

This code reads our transcript now line by line to (1) identify who is speaking and (2) if they are still speaking in the next line. If they do, then this code merges the two lines from the same speaker (if-part of the function). If they don’t, this code adds the new line from the new speaker (else-part of the function). Ultimately, it produces a new transcript for us, our cleaned transcript.

One extra thing we might want to change in our transcript are our interviewer and participant names. Again, this can be done in R, with a basic “Find and Replace” function:

#de-identify transcript
transcript <- sub("Max Musterman", "Interviewer1", cleaned_transcript)   #replace Interviewer Name
transcript <- sub("Erica Musterman", "Participant1", cleaned_transcript) #replace Participant Name

The only thing left to do is to save your cleaned transcript. We want to save it, like before, as a text file (e.g., .txt, .vtt) which can then be further imported into Word, NVivo, Excel, or just simply printed out. Again, we are using the readr package from the tidyverse for this.

#save cleaned transcript
write_lines(cleaned_transcript, "/transcript_cleaned.txt")

Can this be streamlined?

Yes! It can. Because most likely, you’ll not only have just one transcript to deal with but, depending on your recruitment efforts, multiple! Rather than going through each step individually for each transcript, you may want to automate this cleaning pipeline using a reusable function in R.

clean_transcript <- function(file_path,
                         id_map = c("ZoomName" = "Interviewer1",
                                    "ZoomName" = "Participant1")) {

#load transcript
transcript <- readr::read_lines(file_path, skip = 1, skip_empty_rows = TRUE)

#de-identify transcript
for (original in names(id_map)) {
  transcript <- sub(original, id_map[original], transcript)
}

#remove clutter
transcript <- transcript[!grepl("^\\d+$", transcript) & #remove lines that contain only numbers (e.g, 330)
                        !grepl("-->", transcript)]      #remove timestamp lines (e.g., 00:31:56.000 --> 00:31:58.360")

#merge consecutive lines
cleaned_transcript <- c()
current_line <- transcript[1]

for (i in 2:length(transcript)) {
  prev_speaker <- sub(":.*", "", current_line)
  this_speaker <- sub(":.*", "", transcript[i])
 if (prev_speaker == this_speaker) {
  text_only <- sub("^[^:]+:\\s*", "", transcript[i])
  current_line <- paste(current_line, text_only)
 } else {
  cleaned_transcript <- c(cleaned_transcript, current_line)
  current_line <- transcript[i]
 }
}
cleaned_transcript <- c(cleaned_transcript, current_line)

return(cleaned_transcript)
}

The function can then be called upon using the following command:

#using file.choose()
cleaned_transcript <- clean_transcript(file.choose(),
                                       id_map = c("Max" = "Interviewer1",
                                                  "Erica" = "Participant1"))
#specifying file path
path <- "/transcript1.vtt"
cleaned_transcript <- clean_transcript(path,
                                       id_map = c("Max" = "Interviewer1",
                                                  "Erica" = "Participant1"))

That’s all.

You can download the full script via my GitHub page. I haven’t tested this yet with interviews or focus groups with more than two speakers, though technically, if slightly adapted (i.e., id_map) the code should work just fine. Of course, there might be an issue, if the Zoom Name, which is included in the transcript, contains a colon, however, in those instances, the names can be deidentified prior to merging the lines.

Is this useful to you? What else should be added? Let me know!

Writing CRediT statements should be easier!

At the point, where you are in need of a CRediT statement, most likely you’ll already have gone through writing your 8,000 word manuscript, formatting according to the wims and wishes of the journal, and gone back and forth in the GoogleDoc comments section with all your co-authors where and how to place that comma. Or, more likely, you’ll have completely forgotten about it entirely, and are now at the point of submission and are trying to remember who did what and what does “Investigation”, or “Conceptualization” actually mean again?

This little CRediT Author Contribution Tool is something I have been meaning to create for a long time. It was one of those things at the bottom of my to-do list, which never got the proper time and attention. And yes, I did Google and found a few other websites that created something similar. Still, I quite like the structure and design of my tool, and, after all, none of the other sites have the names of my cats and my wife hidden in it.

It’s definitely not earth-shattering. Still, I think it’s one of those gadgets that’s worth bookmarking for future use and may help you down the line.

What are CRediT roles?

The CRediT taxonomy was first developed by Brand et al. (2015). It is a system that serves to identify the contributions of each author on a paper, so that those authors who did most of the heavy lifting, also get the acknowledgment that they deserve. CRedit roles are, for example, Conceptualization (i.e., developing research frameworks, tools, or experimental paradigms), or Investigation (i.e., performing the experiments, or data collection).The full list of roles and explanations can be found on the official CRediT website.

And what does this thing do?

It’s no rocket science (although it may be used in that context). In a two-step process, you (1) enter the names of all the authors on the paper, which you then (2) identify by their CRediT roles. The tool gives you short descriptors of applicable CRediT roles, so that you can make sure, you are actually checking the right boxes. When you are ready, you click “Create CRediT Statement”, and copy and paste the provided output into your document. Also, don’t worry: No data is collected here.

That’s all.

Is this useful? What else should be added? Let me know!

While the published version was slightly edited, this is the originally submitted letter to the editor.

Dear Editor,

We were very interested to read the letter of the day published in your newspaper on January 13^th. The writer, a primary school teacher, insightfully urged a discussion on younger children’s screen use that goes beyond the current policy focus of teenagers and social media. Their letter placed a much required emphasis on screen use in early childhood when the most rapid periods of brain and behavioural development occur. Despite the current shift of focus to AI in terms of global digital technology research, we are still lacking answers to basic questions on the influences and effects of screen use in infants and young children. In our Babylab at University College Dublin we explore environmental influences on children’s early development. One of our primary areas of interest relates to the content and contextual factors of toddler screen use as well as the influence of parents’ own screen use.

In a recent survey of 188 parents across Ireland, we found that 30% of toddlers exceed the HSE-recommended threshold of 60 minutes or less screen use per day. Moreover, we identified a wide range of digital programming that young children regularly engage with across modern streaming services and video platforms (e.g., Netflix, Disney+, YouTube). The current explosion in digital programming for young children may include offerings that might not be designed for their developmental stage. What’s more, the design of streaming platforms, where the end of one episode of a favourite cartoon rolls into the start of the next could make it difficult for parents to track children’s screen time. This is very different to the experience of live TV for previous generations.

The HSE published screen time guidelines for the first time in 2024, yet, beyond duration there is a lack of guidance for parents on content factors such as highly stimulating cartoons (and games) and contextual factors such as social support for parents. In this digital age it is important to provide more nuanced and practical guidance and support for children’s early engagement with screens to ensure optimal developmental outcomes for all children.

Yours sincerely,
Tobias Constien at the UCD Babylab