Friday, 2 October 2015

Using Twitter data to provide insights into health conditions and health-related events (by Wasim Ahmed)

My research examines social media data, such as data derived from Twitter, to provide insights into health conditions and health related events.

Twitter has 316 million monthly active users and there are 500 million tweets per day.  It can be used as a source of data for social science research both current and historical in and of itself, but it can also be used to complement more traditional data sources, such as surveys and interviews.  I lead the New Social Media New Social Science Network (NSMNSS) Twitter account, which has  members from across academia and industry who explore the methodological implications of social media research.

One of my case studies focuses on the Ebola outbreak of 2014, where I have amassed at least 26 million tweets. Examining tweets allows the real-time monitoring of public views and opinions. These can be monitored by people from the health sector who can then disseminate accurate information appropriately.  In some instances, data derived from Twitter allows geographical surveillance, and has the potential to be used to identify locations of possible infectious disease outbreaks.  Twitter has proved useful in emergency and crisis situations.

There are often specific methodological, ethical, privacy, and copyright issues which require careful consideration, and my PhD research also critically considers these.  I am also aiming to identify and evaluate the software that can be used by social scientists or those from the health sector, to analyse Twitter. This is very important, as it allows non-computer scientists or non-programmers to retrieve Twitter data in order to ask social science research questions.

Since the start of my PhD I have been disseminating thoughts and findings.  I am an active tweeter, and my research blog has proven to be very popular.  Some posts have appeared in Google Scholar and others have been picked up by the mainstream media.  My research has been mentioned on the British Medical Journal’s (BMJ) blog, ihawkes blog, DiscoverText blogs (for a historical data prize).  I receive regular invitations to academic and industry events and have recorded an audio lecture for a group of Masters students at Western Sydney University on how to retrieve data from Twitter, and on the methodological implications of social media research.

Tuesday, 18 August 2015

What is Chemical Similarity, and how is it Useful? (by Edmund Duesbury)

In the final year of my PhD, I have been investigating different forms of alignment of chemicals, and seeing which method is best at predicting whether a chemical will be active against a particular drugtarget.

Similarity is subjective to a particular problem domain.  As an example, which two most objects are most similar – an apple, a pumpkin or a basketball?  All three are more or less spherical, but the pumpkin and apple have the similarity of being fruit, while the pumpkin and basketball are a similar size.

The same subjectivity exists in chemistry.  A common goal when searching for similarity in chemicals is to predict whether one compound will act in the same way as another compound, known to have useful pharmaceutical properties.  The desired “similarity” in this case, is a similarity of biological activity: something which, at present, is impossible to predict.  However, we can attempt to infer such a property from aspects of structural similarity.

Serotonin reuptake inhibitors is a group of chemicals that includes many useful antidepressants. Consider the examples below, of compounds that act as serotonin reuptake inhibitors.  In the first case (Figure 1), similarity is based on the largest common fragment (highlighted in bold).
Figure 1

The similarity here is obvious, the only difference being the Br atom.  However, the same technique fails to show the biological similarity between the two inhibitors in Figure 2.

Figure 2.

In this case, the approach of finding the largest fragment has failed to highlight the “similarity” between the two compounds.  A technique based on finding the maximum possible overlap of edges however, is more successful (Figure 3).

Figure 3.

This method, which seeks to find a set of common fragments emphasises a different “similarity” between these compounds.

Wednesday, 12 August 2015

Perish by peer review

Many thanks to Sheila Webber for her entry on this blog regarding the recent Times Higher Article (THE) on peer review.  The article is well worth reading and makes several valid points.

However, a distinction should be made between peer review and anonymous peer review.  Peer review (both formally and informally) is a widespread practice in many professions (including academia).  Anonymous peer review used to be the preserve of academics, but (thanks to Web 2.0) is now a feature of 21st century life.  Anyone who has selected a restaurant on TripAdvisor is likely to be reading an anonymous review by a peer from the dining community.

One key difference is that a bad review on TripAdvisor can be countered by the business owner, and is not likely (on its own) to bring down the business.  The nascent careers of academics are more fragile.

Another key difference is that, on TripAdvisor, I can visit the restaurant and read the reviews, then assess how representative the review is of my interests and my tastes.  In other words, how much of a peer was the reviewer?

When academic peer review began, it was not anonymous.  At some time after WWII, arrangements by which journal editors informally contacted academics for advice on submitted articles appear to have been formalised in the system of anonymous peer review.  Prior to that, the interests, characters and prejudices of reviewers would have been known, resulting in greater openness and (occasionally) unrestrained unpleasantness.  However, the reviewer's credentials could be assessed and (if necessary) questioned.

One contributor to the THE article notes that, in his discipline alone (economics), there are 20,000 new journal articles every year.  The pool of reviewers must therefore be very large, prompting the question: to what extent is a reviewer a peer of the author?

The THE article ends with a contribution from an anonymous author who is is establishing a website for particularly bad examples of anonymous peer review.  I would certainly applaud such an exercise but I hope that, as well as giving the opportunity to read the poor reviews, the site also publishes the articles that attracted them.  It will be interesting to see how many suffer from poor writing, poor research and poor analysis, and how many suffer from being unorthodox and innovative.

Friday, 7 August 2015

Peer review or perish

A couple of years ago, Andrew wrote an interesting piece on peer review, which you can find here Published yesterday in the Times Higher is a piece in which 6 academics (mostly full professors, I note) quote the worst peer reviews they ever got, and give their opinion on whether peer review should be jettisoned. The reviewer comment “What is this muck?” is one of the most arresting (in response to a paper by Prof Susan Bassnett). The last contributor chooses to remain anonymous, but is calling for examples of bad peer review to put on a website...
Times Higher Education. (2015, 6 August). The worst piece of peer review I’ve ever received. Times Higher Education.

Thursday, 16 July 2015

Maths, nuclear explosions and other recreational activities

It's always nice to get a comment from someone who isn't trying to sell something.  Zoe Chen's comment not only displayed cultural awareness (math / maths) but also touched on points that were raised during the discussion with Marc.

The three games Marc is hoping to get feedback on are:
Ordering Fractions Game and Giving Change Game by the BBC and
Algebra Meltdown by Manga High.

"What's that last one about?" someone asked
"It teaches maths by getting people to solve algebraic equations in order to prevent a nuclear explosion" answered Marc.
"Er... so people who are nervous about maths are told 'If you get these equations wrong, there'll be a nuclear disaster.'  That's very homoeopathic!"

Marc made the point that storytelling could be a factor in the success of a game; and actually the Algebra Meltdown game (despite its name) isn't that scary.  The only things that explode (metaphorically) in the event of algebraic error are nuclear scientists.  I suppose it's difficult to get the balance of punishment and reward right in such things.  If you make the game too stressful, students give up on maths and seek therapy*.  However, if the consequences of making a mistake are comical, students deliberately make mistakes.

*I met a New York based therapist recently.  He charges $200 per hour.  I guess they rely on people not being able to add up. 

Monday, 15 June 2015

Title: Educational computer games and their impact on maths anxiety in University social science students (by Marc Bonne)

Educational computer games are already known to help motivate university students to learn maths. However, few studies have measured the impact of such games on maths anxiety: a significant factor contributing to underperformance in maths-based studies.

This project aims to assess the degree to which educational computer games affect maths anxiety in university social science students. Maths anxiety is prevalent amongst such students because many lack a maths-oriented background, but are often being exposed to statistics and other maths subjects during their courses.

The study consists of a systematic literature review of educational games and maths anxiety (currently underway).  When this is completed, students from the Faculty of Social Science in the University of Sheffield will be invited to complete the Maths Anxiety Rating Scale (MARS) brief scale in order to assess the scale of their maths anxiety. Participants will then put into a control group and an experimental group. The control group will carry on their course as normal, while the experimental group will play an online maths game.

The experimental group will be observed as they play the game.  Data collected will include think-aloud data; records of keyboard and mouse strokes; and video recordings of gameplay and body language.

A sample of participants from the experimental group will be interviewed in order to gauge how they felt while playing the game.  This, in conjunction with the video data, will help identify which aspects of the game provoke an emotional response .

A post-session maths anxiety scale will be given to both the control group and the experimental group in order to measure any difference in maths anxiety levels before and after the session. Furthermore, the average maths anxiety levels of each group are to be compared and analysed.

Several online maths games have been identified as possible choices to be used for data collection.  A pre-pilot study is to take place where potential participants play each of the three online games and decide which one to use for the main study based on the games’ usability.

Following the pre-pilot study, a pilot study will take place involving the same procedures as the main study but with smaller sample sizes.

Tuesday, 19 May 2015

An exploration of the information literacy experiences of ESOL (English for Speakers of Other Languages) learners (by Jess Elmore)

My research explores information literacy in the context of the ESOL classroom.  I am interested in finding out about the relationship between language learning and information literacy, and about how changes to learners' information literacy practice impact on their lives.  My working definition of information literacy is that it is a sociocultural practice; a set of shared activities rather than just individual skills that constitutes the ability to find, use and share information in a particular context or information landscape.

ESOL learners are typically people who have come to the UK for work, for family reasons, or to claim asylum; and are learning English as part of adult basic skills provision. They are a very diverse group but can be seen as disadvantaged by several measures; they are immigrants, they are less likely to be employed, they are generally female, and they often come from BME (black, minority and ethnic) communities. Information literacy and ESOL can be seen to have similar goals in terms of providing individuals with the ability to participate fully in society (however you choose to interpret this) but the relationship between the two has not been explored in detail.

I am planning a longitudinal case study of three community ESOL classes. My research is multi-method; I will use observation, one to one interviews, focus groups, group interviews and visual methods. My research is participatory and emergent so I will negotiate with participants what methods to use and hope to involve them in the research as far as possible. The multi-method approach is used because I am interested in rich, holistic information experiences, but also to help overcome the language barriers present when working with participants who have limited English.

I have completed a pilot study consisting of one observation of an ESOL and Art workshop and two focus groups which were held in existing ESOL conversation classes.  The findings from the pilot were local and limited, but suggested several areas for further exploration: in particular, the diversity of ESOL learners' information experiences, their use of digital technologies, the significance of religion, and the importance of place and people.  However the methodological findings were more interesting. The pilot identified the language level of ESOL learners who could talk meaningfully about their information experiences, but also raised important questions about my position as a researcher, the process of analysis and the need for sustained research relationships rather than single encounters.