Month: April 2015

How many times is “Science” mentioned in each party’s manifesto?

The U.K. General Election is fast approaching, and each political party is clamouring for the public’s attention. All main parties have now published their manifestos: the document outlining plans and policies the party will implement if in office. In them you will find promises of reducing unemployment, cutting the nation’s deficit, and improving the National Health Service. 

But what about science? I was interested in whether the parties had mentioned science at all in their manifestos. After all, each manifesto page is prime real-estate for publicly-popular policies (hey, how else can the party get nominated?). So, I downloaded each party’s manifesto, and did a search for the occurrence of the word “science”. The results are plotted below, with some caveats:

  1. it was merely a count of the word “science” appearing in each manifesto—I made no judgement about what was being discussed (i.e., it could have said “We will never invest in science”, and would still receive a count of one
  2. The number of pages (i.e., the scope of the real-estate) varied between parties, so there is a second plot which shows the count as a proportion of total page numbers in the manifesto.

R Code for Reproducing the Plots

# clear all
rm(list = ls())

# you need the ggplot2 package for this

# declare parties
Party <- c("Conservative", "Labour", "Lib. Dems.", "UKIP", "SNP",
 "Plaid Cymru", "Green")
Party <- factor(Party, levels = Party)

# count for the word "science"
Count <- c(16, 3, 7, 6, 1, 5, 10)

# pass all to a data frame
allData <- data.frame(Party, Count)

## do the plotting for raw counts of the word "Science"

# define party colours
# (see
cols <- c("deepskyblue", "firebrick1", "gold", "purple", "goldenrod", "green4",

# now plot
p <- ggplot(data = allData, aes(x = Party, y = Count, fill = Party))
p <- p + geom_bar(stat = "identity")
p <- p + scale_fill_manual(values = cols)
# p <- p + coord_polar(theta = "y") # un-comment if you want a pie-chart
p <- p + guides(fill = FALSE)

## do the plotting of the proportion of the occurence "Science" to the number
# of pages in the manifesto

# how many pages in each party's manifesto?
nPages <- c(84, 86, 158, 76, 56, 64, 84)

# what is the proportion?
Proportion <- Count/nPages

# remove NAs
Proportion[is.nan(Proportion)] <- 0

# now plot
p <- ggplot(data = allData, aes(x = Party, y = Proportion, fill = Party))
p <- p + geom_bar(stat = "identity")
p <- p + scale_fill_manual(values = cols)
# p <- p + coord_polar(theta = "y") # un-comment if you want a pie-chart
p <- p + guides(fill = FALSE)


Grant-Givers: Embrace (and Fund!) Research Without Impact

impact / n.  the benefit or contribution to society of research

wordle 2

Researchers in the U.K.—and likely elsewhere—will be no strangers to the term “impact”. Our research is supposed to have it. We are supposed to evidence it. Our institutions in their Research Excellence Framework—the “system for assessing the quality of research in UK higher education institutions”—have to provide case studies of it. But what is “impact”? Why is it deemed important? More importantly, is there room for research without impact? I argue that yes, there is room, and grant-givers should be embracing (and funding!) it.

What is Impact, & Why is it Important?

As the definition at the head of this post states, impact is generally considered to be the benefit or contribution to society of research, a definition echoed by the Economic and Social Research Council (ESRC)—Britain’s premier funder of social-science research. Importantly in the context of this post, impact is generally considered to be non-academic; just because your recent research was published in Nature, it doesn’t automatically qualify as impact.

The ESRC state—on their excellent online resources for researchers wishing to submit a grant to them—that evidence is essential to qualify something as impact, for example “…that it has been taken up and used by policy-makers, and practitioners, has led to improvement in services or business”. Impact can be generated via many avenues: via public engagement & knowledge exchange programmes, networking with stakeholders and participatory groups, direct engagement with end-users and/or practitioners etc. 

But why is impact important? It might be abundantly clear why impact might be important to university managers, as research conducted at their institutions that has led to serious impact can only boost the university’s prestige (which in turn leads to student numbers, which in turn leads to more income, which in turn….). It is also clear why the government—who typically fund a lot of research—want impact; they paid for it with taxpayers’ money, and lord knows the politicians don’t want to be seen wasting taxpayers’ money.

If research is being conducted to contribute meaningfully to society and/or the economy, then society itself should also value impact. This can ensure that decisions being made by policy-makers are based on research evidence, which can only be a good thing. Researchers also should value impact; who among us doesn’t want to contribute to society in some small way? We all want our research to be meaningful.

What’s the Problem with Impact?

Given all of the above, you would be forgiven for thinking what my problem with impact is. It boils down to the distinction between basic and applied  research. Basic research (also known as pure/fundamental research) is research conducted without a practical outcome in mind; that is, it is the pursuit of knowledge for its own sake. Applied research applies knowledge and the scientific method to a particular problem or a particular purpose. 

I am a basic psychological scientist. I do no applied research. I do not have applied questions in mind when conducting my research. I aim to understand a small portion of cognition: cognitive control processes. That is my aim, and only this. To understand. My problem with impact is that it is heavily biased towards applied research. This wouldn’t be problematic, necessarily, were it not for the fact that many grant-giving agencies in the U.K. require a comprehensive plan for how the research currently being proposed will lead to demonstrable—non-academic—impact. As a consequence, an application geared solely to a basic research question potentially lends itself to less impact than an application addressing an applied research question. There seems to be an imbalance. Should we prioritise research that is applied in nature? I argue NO. 

I am not the first to argue that shunning basic research because it has no immediate application is short-sighted. As Nobel prize-winning chemist Sir George Porter wrote: “To feed applied science by starving basic science is like economising on the foundations of a building so that it may be built higher. It is only a matter of time before the whole edifice crumbles.” Basic research is the foundation upon which applied research can be built. The former is important in its own right, but the latter is difficult (impossible?) without the former. 

Basic research develops human understanding, a noble pursuit in its own right. An added benefit is that with this increased understanding, applications can arise naturally. Many technologies in use today arose from basic research, without its current use in mind. My favourite example is the origin of all of our digital world: binary coding. Wikipedia states that binary code was developed by Indian scholar Pingala in the 2nd century BC. Surely this chap didn’t have digital communication via iPhones in mind when developing this method!

A problem with pursuing only applied research is that our understanding—which can only be developed via basic research—stagnates. We need the basic research to push the applied research. My favourite quote regarding this point comes from another Nobel laureate—George Smoot—who stated: “People cannot foresee the future well enough to predict what’s going to develop from basic research. If we only did applied research, we would still be making better spears“. 

Grant-Givers: Embrace (and Fund!) Research Without Impact

There exists a problem in psychological science: careers can be built on the grant income a researcher can bring to her department. A budding career means more time to devote to the research. But this raises an immediate problem: If impact is the game, and basic research has little/no impact, basic research won’t ever be a priority for funding. Therefore, those pursuing basic research will likely jump ship to the applied-research game in order to advance their careers. But as I’ve discussed, this means that psychological science will be “economising its foundations”. 

Therefore, grant-giving agencies should develop programs purely directed to basic research in psychological science. Let’s first build a strong foundation of understanding before building the skyscraper of application.