Several weeks ago, I posted an analysis of tweets about the restrictions imposed on Melbourne residents in an effort to control an outbreak of Covid-19. That analysis was essentially a road-test of a Knime workflow that I had been piecing together for some time, but that was not quite ready to share. Since writing that post, I have revised and tidied up the workflow so that anyone can use it, and I have made it available on the Knime Hub.
In the present post, I provide a thorough description of the workflow, which I have named the TweetKollidR, and demonstrate its use through a case study of yet another dataset of tweets about Melbourne’s lockdown (which, as I write this, still has not ended, although it has been eased). 1
As you will see from the search queries in Figure 3, this dataset includes some keywords that relate to Victoria more generally, rather than just Melbourne. However, since most of the content concerns the Melbourne lockdown, I will continue to refer to it as such. ↩
It’s been part of my routine for several weeks now. Almost like clockwork, at around 8:30am, my phone buzzes. I hold my breath, partly avert my gaze, and unfold the notification just far enough to see the familiar sentence: “Victoria has recorded…”
What comes next can set the tone for the whole day. If the number of new Covid-19 cases recorded in the preceding 24 hours is smaller than the number reported the day before, I breathe a sigh of relief and ease into the day as if a small weight has lifted. If the number has gone up, I slap down my phone in disappointment and try — usually without success — to console myself with the idea that it is probably just a blip that will be corrected with a lower number tomorrow.
I’m sure that the story is similar for many Victorians. It could hardly be otherwise, given that these daily case numbers are now our ticket to freedom, as laid out in the state’s Roadmap to Covid-normality. If the case numbers stay low until September 28 — or more specifically, if the 14-day average at that point is less than 50 — Dan Andrews will let Melbourne residents socialise with up to five people from two households. Luxury! If we get the average down to below five by 26 October, we’ll almost be allowed to behave like human beings again. Under the current framework, our only way out of lockdown is through the numbers.
Given that the only numbers that really matter according to the roadmap are fortnightly averages, it makes little sense to get worked up about the number of cases announced on any given day. Probably we’d all be better off ignoring the daily announcements and getting weekly summaries instead. But I, for one, am not about to kick the habit. As long as that magic number is reported each day, I am going to keep getting my fix and reading into it as much as I can.
Weekly Covid cycles
One thing that has become apparent to case number junkies like me is that not every day of the week is equal. On average, certain days of the week tend tend to have higher case numbers than others. You can see this most clearly in the global total, as in the version reproduced below from Our World in Data. Nested within the larger wave of cases is a recurring ripple of a week’s duration.
Surprisingly, there is no consensus yet about why this weekly cycle occurs. The handful of research papers that I have found about the topic all confirm that weekly cycles in cases and deaths are real, but offer contrasting explanations. One paper examining data specific to the US concluded that most of the weekly variations could be explained by quirks in reporting regimes and fluctuations in testing activity. Other studies, especially those looking at countries other than the US, have argued against this explanation and suggested alternative causes. Noting that new cases in several countries tend to peak on Thursdays or Fridays and then fall on weekends, one paper hypothesises that infections rise when the stress of the working week compromises the immune system. Another paper explains the same pattern by suggesting that weekends provide more opportunities for young people to mingle with their elders, thus causing infections that will become symptomatic five days later (i.e. on Thursday or Friday) and leading to deaths about 14 days after symptoms emerge. Yet another paper hypothesises that cycles in air pollution (caused by traffic, for example) or the bodies own circadian rhythms could play a role.
This kind of weekly oscillation has not been as obviously apparent in Australian case numbers, largely because the numbers have been so low to begin with. When cases did get out of control in Victoria a couple of months ago, the Stage 4 lockdown measures introduced in early August turned the numbers around so quickly that there has never been a stable baseline against which to notice more nuanced levels of variation. Even so, I’ve noticed on several occasions that the numbers reported on a Monday are relatively low; and I recently heard Casey Briggs refer to ‘hump day’ in one his regular case reports on the ABC (even if I didn’t catch which day he was actually referring to). I’ve also heard vague references by media commentators to backlogs and fluctuations in the processing of test results, which could influence the number of cases announced on any given day.
This is the sort of information that you need to know if, against your own better judgement, you are going to try to extract some kind of meaning from the daily announcement of new case numbers. As I write this, tomorrow is Friday. If the number that pops up on my phone just after breakfast is hardly any lower than today’s, how worried or surprised or disappointed should I be? Is Friday a day when the numbers tend to be higher or lower than would be dictated by the underlying trend?
I couldn’t find any existing answers to this question, so I got hold of Victoria’s daily case data and took a stab at answering it myself. I should stress that I did this by following my own statistical intuitions rather than emulating any of the methods used in the papers mentioned above (most of which I hadn’t read until after I did this!). I think my approach makes sense, but I make no claims to it being the best method available. If it turns out that I’ve committed some kind of crime against statistics here, I’ll humbly (indeed gratefully) accept a fine from the statistics police. Continue reading Is there a weekly cycle in Victoria’s Covid case numbers?→
What we talk about when we talk about the lockdown
Back in January, I wrote a lengthy, data-driven meditation on the merits of my relocation from Brisbane to Melbourne. My concern at that time was the changing climate. Australia had been torched and scarred by months of bushfires, and I was feeling pretty good about escaping Brisbane’s worsening heat for Melbourne’s occasionally manic but mostly mild climatic regime.
But by gosh do I wish I was back in Brisbane now, and not just because Melbourne’s winter can be dreary. While Brisbanites are currently soaking up as much of their famed sunshine as they like, whether on the beach or in the courtyard of their favourite pub, Melburnians are confined to their homes, allowed out of the house for just an hour a day. During that hour, we are unable to venture more than 5km from our homes or to come within 1.5 meters of each other, leaving little else to do but walk the deserted streets and despair at all of the shuttered bars, restaurants and stores. All in the name of containing yet another existential threat that we can’t even see.
Of course, just because we can’t see the coronavirus doesn’t mean we can’t talk about it. Indeed, one unfortunate consequence of the ‘Stage 4’ lockdown 1 that’s been in place in Melbourne since the 2nd of August is that there is little else to talk about. We distract ourselves from talking about how bad things are by talking instead about how things got so bad in the first place. On days when our tireless premier (who at the time of writing has delivered a press conference every day for 50 days running) announces a fall in case numbers, we dare to talk about when things might not be so bad any more.
To date, we’ve been from Stage 3 back to Stage 2, and then up again to Stage 3 before ratcheting up to Stage 4. Hopefully we’ll be back to Stage 3 in a few weeks. We keep using that word, but I don’t think it means what we think it means. If I lapse into calling it ‘Level 4’ instead, that’s why. ↩
As documented elsewhere on this blog, I recently spent four years of my life playing with computational methods for analysing text, hoping to advance, in some small way, the use of such methods within social science. Along the way, I became interested in using topic models and related techniques to assist the development of public policy. Governments regularly invite public comment on things like policy proposals, impact assessments, and inquiries into controversial issues. Sometimes, the public’s response can be overwhelming, flooding a government department or parliamentary office with hundreds or thousands of submissions, all of which the government is obliged to somehow ‘consider’.
Not having been directly at the receiving end of this process, I’m not entirely sure how the teams responsible go about ‘considering’ thousands of public submissions. But this task strikes me as an excellent use-case for computational techniques that, with minimal supervision, can reveal thematic structures within large collections of texts. I’m not suggesting that we can delegate to computers the task of reading public submissions: that would be wrong even if it were possible. What we can do, however, is use computers to assist the process of navigating, interpreting and organising an overwhelming number of submissions.
A few years back, I helped a panellist on the Northern Territory’s Scientific Inquiry into Hydraulic Fracturing to analyse concerns about social impacts expressed in more than 600 public submissions. Rather than manually reading every submission to see which ones were relevant, I used a computational technique called probabilistic topic modelling to automatically index the submissions according to the topics they discussed. I was then able to focus my attention on those submissions that discussed social impacts, making the job a whole lot easier than it otherwise would have been. In addition, the topic model helped me to categorise the submissions according to the types of social impacts they discussed, and provided a direct measurement of how much attention each type of impact had received.
This experience proved that computational text analysis methods can indeed be useful for assessing public input to policy processes. However, it was far from perfect case study, as I was operating only on the periphery of the assessment process. The value of computational methods could be even greater if they were incorporated into the process from the outset. In that case, for example, I could have indexed the submissions against topics besides social impacts. As well as making life easier for the panellists responsible for other topics, a more complete topical index would have enabled an easy analysis of which issues were of most interest to each category of stakeholder, or to all submitters taken together.
In this post, I want to illustrate how topic modelling and other computational text analysis methods can contribute to the assessment of public submissions to policy issues. I do this by performing a high-level analysis of submissions to the Victorian parliament about a proposal to expand Melbourne’s ‘free tram zone’. I chose this particular inquiry because it has not yet concluded (submissions have closed, but the report is not due until December) and because it received more than 400 hundred submissions, which although perhaps not an overwhelming number, is surely enough to create a sense of foreboding in the person who has to read them all.
This analysis is meant to be demonstrative rather than definitive. The methods I’ve used are experimental and could be refined. More importantly, these methods are not supposed to stand on their own, but rather should be integrated into the rest of the analytical process, which obviously I am not doing, since I do not work for the Victorian Government. In other words, my aim here is not to provide an authoritative take on the content of the submissions, but to demonstrate how certain computational methods could assist the task of analysing these submissions. Continue reading Free as in trams: using text analytics to analyse public submissions→