Content uploaded by Emma Weitkamp
Author content
All content in this area was uploaded by Emma Weitkamp on Jan 07, 2014
Content may be subject to copyright.
Article
Corresponding author:
Emma Weitkamp, Science Communication Unit, Faculty of Health and Life Sciences, The University of the
West of England, Bristol BS16 1QY, UK.
Email: emma.weitkamp@uwe.ac.uk
Podologues: conversations
created by science podcasts
Hayley Birch
Emma Weitkamp
Science Communication Unit, The University of the West of England, UK
Abstract
Podcasts are media files that can be automatically aggregated and downloaded via the
internet, and transferred to portable media players. Combined with online discussion
facilities, podcasts represent flexible and potentially valuable tools for communicating
about science. This pilot project aimed to assess the role of science podcasts in
stimulating discussions, or ‘podologues’, about science through detailed analyses of a
sample of five popular science podcasts. Two main methods were used: content analysis
of online discussion forums and blogs associated with the five podcasts and interviews
with listeners. The results show that podcasts are regarded as valuable sources of
scientific information by listeners and that blogs and forums can act as public spaces
for audience members to share knowledge, develop their own ideas about science and
provide feedback to media producers. Larger, more detailed studies are required to
further understand the value of podcasts for stimulating public discourse about science.
Key words
interactivity, online forums, podcasts, science communication
Introduction
Podcasts and the evolving media landscape
Podcasting has emerged alongside large-scale changes within the media. Figures for
traditional print media and television consumption map a downwards trend, while inter-
net audiences continue to soar (Eggington, 2008). Radio audiences, on the other hand,
new media & society
12(6) 889–909
© The Author(s) 2010
Reprints and permission: sagepub.
co.uk/journalsPermissions.nav
DOI: 10.1177/1461444809356333
http://nms.sagepub.com
890 new media & society 12(6)
have been rather less affected by the rise of the internet (Merzagora, 2004; RAB/IAB,
2006) and the popularity of online ‘listen again’ facilities (Berry, 2006; RAJAR/Ipsos
MORI, 2008) proves that there is still a great demand for audio content. Initial optimism
about the future of digital radio, however, has waned (Griffiths, 2008), suggesting that
the advantages over traditional radio are either not being made clear enough, or not liv-
ing up to expectations. Listen again services, online radio and podcasts, by contrast, offer
very clear benefits. Consumers can choose from a very wide range of content and, per-
haps most importantly, control their own listening schedules – a recent UK survey found
that three quarters of podcast listeners are listening to episodes more than a week old
(RAJAR/Ipsos MORI, 2008). The desire for control of schedules for personal media use
was also highlighted in a survey of ‘next generation’ radio listeners (The Knowledge
Agency, 2004).
Online consumers are being given more control in other ways. The rise of blogs,
comment facilities, forums and collaboratively edited sites – all under the guise of ‘Web
2.0’ – is giving listeners and readers opportunities to generate their own media content
(Minol et al., 2007). In this sense, traditional models for the way media information is
distributed are disintegrating. Members of the public can now see their comments pub-
lished on the websites of major newspapers. And whether or not this diminishes the role
of the professional journalist, as some would argue (DeLong, 2007; Keen, 2007), it
clearly influences the media agenda.
Conversations in the ‘global’ online community
McLuhan (1962) predicted that the media would cause the world to become a ‘global
village’ where media consumers could experience distant events as though they were
there. It could be argued that the internet has made the world an even smaller ‘village’,
allowing users to interact with each other in the public domain. Through the internet, two
geographically disparate individuals can become close within the new kind of ‘local
community’ (Postill, 2008) – a community that is based around shared web space and
virtual vicinity to each other. Where communities are based around shared reading or
listening habits, there is a clear focus for conversation. In discussing particular podcast
episodes on forums, for example, users might improve their understanding and expand
their knowledge of topics covered.
Minol et al. (2007) claim that new media tools such as forums, blogs and wikis pro-
duce a surplus of irrelevant and unreliable information. With regard to science content,
they say the media should tap the scientific source. On the other hand, Jenkins (2006)
argues that web communities are very successful at filtering out aberrant information,
citing the online encyclopaedia Wikipedia as an example of publicly authored webcon-
tent that is, on the whole, fairly accurate. Such enterprises rely on the vigilance and
trustworthiness of the community as a whole to ensure that rogue comments are deleted.
In the same way, users of a forum can be expected to ‘boot out’ those who break the com-
munity’s moral code.
In 1997, collective intelligence theorist Pierre Lévy envisaged the emergence of a
global virtual knowledge space. He saw the expanding internet as a place for building
communities and exchanging information.
Birch and Weitkamp 891
Internetworked data would … provide the technical infrastructure for the collective brain or
hypercortex of living communities. The role of information technology and digital communica-
tions is to … promote the construction of intelligent communities in which our social and
cognitive potential can be mutually developed and enhanced. (Lévy, 1997: 9, emphasis in original)
Most sources would agree that the main function of Web 2.0 is to facilitate sharing of
information and content (McDermott, 2008; Schmitz-Justen and Wilhelm, 2007), argu-
ably in the way that Lévy imagined in the early days of the internet. Exploring internet-
based discussion forums, Schmitz-Justen and Wilhelm (2007) found that those forums
which focused on knowledge exchange and minimised social discussion were more likely
to elicit contributions that could lead to knowledge sharing and transfer; however, the
initial framing of the discussion played a key role in ensuring contributions remained
focused on the core subject. Thus, if they are well managed, blogs and discussion forums
are immensely powerful tools for communication, allowing individuals from all over the
globe to pool their expertise as shared knowledge within online communities. Nevertheless
recent research suggests web users are, in general, reluctant to contribute to online discus-
sions (Thurman, 2008).
Beaudoin (2008: 469) recently described the internet as ‘ushering forth a closely
intertwined global village’. However, there is a difference between information that is
theoretically available to everyone, and information that everyone can and will access.
First, not everyone can access the internet and this is not just a consequence of lacking
physical or economic resources (Newholm, 2008). Other social and psychological fac-
tors, such as real or perceived lack of technical support, often play a part. Secondly, what
will be accessed depends on the interests of the individual web user. In this context, the
editor of Wired magazine contrasted the different marketing models exploited by traditional
and online media (Anderson, 2004). In coining the phrase ‘the Long Tail’, he described the
endless choice offered by online media at the less mainstream end of the market.
Collaborative filtering … uses the browsing and purchasing patterns of users to guide those
who follow them … This is the difference between push and pull, between broadcast and
personalized taste. Long Tail business can treat consumers as individuals, offering mass
customization as an alternative to mass-market fare. (Anderson, 2004: 5)
Podcasting, as a medium, can be considered a hybrid of ‘push’ and ‘pull’. RSS feeds
‘push’ the content, but it may take even a proactive internet user considerable effort to
seek out and sign up to the feeds they want to receive in the first place. In this sense,
podcasting is most definitely narrowcasting as opposed to broadcasting. If, as Anderson
claims, most people’s tastes begin to depart from the mainstream online, the picture of
the internet as a global online ‘village’ is highly inaccurate. It might be regarded rather
as being made up of millions of tiny hamlets. This suggests that blogs and forums asso-
ciated with science podcasts will attract only modest audiences – and even fewer
contributors.
The problem innate to narrowcasting, and therefore podcasting by some definitions
(Holtz, 2005; Leigh, 2005), is that it caters for niche audiences. From the listener’s point
of view, this can only be described as a benefit – podcasts to suit any interest, or interest
892 new media & society 12(6)
level, can be downloaded, almost always at no charge to the listener. The iTunes pod-
catcher and Podbean and Odeo podcast aggregators all offer hundreds of different pod-
cast feeds just within the Science and Medicine categories, with specific subject matters
ranging from stem cells to particle colliders. There is very little research to indicate who
listens to podcasts (RAJAR/Ipsos MORI, 2008) but one could certainly speculate that in
general they attract people who are already somewhat interested in the subjects covered
in the podcasts to which they subscribe. An evaluation of Astronomy Cast, for example,
shows that listeners have a very strong bias towards science and technology podcasts, as
opposed to news, culture and other topics (Gay et al., 2008).
Podcasting as a tool for science engagement
An appreciation of how science fits into society is also vital to any informed debate about
science (European Commission, 2006) and there has been a well recognised movement
in the science communication field to increase debate and discussion about science. In
the UK, for example, since the publication of the House of Lords’ Science and Society
report (2000), there has been a move away from one-way transmission models of com-
munication to approaches encompassing dialogue and discussion.
Traditional media, such as newspapers and radio, are dominated by a one-way flow
of information from scientists and media producers to readers or listeners. This offers
little opportunity for the listener or reader to interact with the journalist, let alone the
interviewees. Internet-based media, such as podcasts, could create more opportunities
for two- or even multi-way communication. McDermott (2008) suggests further that
podcasts, and their associated blogs and forums, as ‘social media’ encourage feedback
and questions from listeners. Producers themselves may access the blogs and forums,
but listeners can also exchange ideas with other listeners, some of whom may be
scientific experts. Thus, new media, including podcasts, have the potential to bridge
the divide between the unidirectional traditional media and face-to-face dialogue
approaches, which may be limited by their audience reach. Through the use of facilities
such as internet forums, these web-based media could stimulate debate and dialogue not
only between listeners and media producers, but more widely by providing a facility for
listeners to engage in discussion with other listeners, be they scientists or otherwise.
Podcasts, blogs, wiki and social networks in general are opening up new room for debate and
knowledge and, at the same time, they feed a great change in the social and political debate.
(Picardi and Regina, 2008: 3)
Perhaps the most obvious reason that people use any media is for entertainment and
information (Green et al., 2004; Vorderer et al., 2004). Audio as a format is considered a
less formal, more engaging way of communicating about science than written material,
and can be used to establish a more intimate connection with an audience (Merzagora,
2004; RAB/IAB, 2006). Podcasts also offer the additional advantage over radio of com-
plete listener control, meaning that not only are listeners engaged, they are able to choose
to listen when and where they please and replay segments of audio that they have missed
Birch and Weitkamp 893
or want to hear again. The potential of podcasts could lie not just in informing or enter-
taining lay audiences but in stimulating meaningful conversation, for example, facilitat-
ing dialogue between lay listeners and those with a range of expertise, including both
formal and informal knowledge.
Research objectives
This pilot project sought to analyse the potential role of science podcasts in stimulating
discussion about science. Specifically, it sought to:
• Analyse the extent and content of discussions on blogs or discussion forums asso-
ciated with five science podcasts over six weeks.
• Determine how discussions about the selected podcasts serve to enhance the
listening experience.
• Assess the relevance and value for listeners of discussions stimulated by the
selected podcasts.
Methods
Podcast selection
Apple’s iTunes is widely accepted to be the dominant podcatcher (Friess, 2006; Kiss,
2008; McLaughlin, 2006) and therefore podcasts were selected from this source. Podcast
producers have noted that iTunes’ Top 100 ranks podcasts by the number of new sub-
scribers and is therefore often highly variable (Frederick, 2008; Green, 2008). Therefore,
to eliminate the effects of short-lived marketing successes, the Science and Medicine
section was monitored for six weeks. Those in the top 100 for at least four of the six
weeks made up the starting sample of 94 podcasts.
A funnelling strategy was applied for selection of the final podcasts (and their associ-
ated integrated online discussion forums [IODFs]). Inclusion criteria were: presence of
an IODF with posts for at least four of the six weeks; and that the podcast should be the
highest ranked podcast from its publisher. Podcasts were excluded if they covered mainly
pseudoscientific or non-scientific, as opposed to scientific, topics. In the context of this
research, science was defined as including the natural sciences and mathematics. Podcasts
focusing on philosophy, scepticism and the paranormal were excluded. The final five
were obtained by eliminating the lowest ranked according to the iTunes ratings and were:
• Astronomy Cast (http://www.bautforum.com/astronomy-cast/ and http://www.
astronomycast.com/)
• The Naked Scientists (http://www.thenakedscientists.com/forum/index.php?
board=16.0)
• WNYC’s Radio Lab (http://blogs.wnyc.org/radiolab/category/podcasts/)
• Guardian Science Weekly (http://blogs.guardian.co.uk/science/category/podcast_1/)
• The Math Factor (http://mathfactor.uark.edu/)
894 new media & society 12(6)
Podcast analysis
All episodes published within a randomly chosen monitoring period of six weeks (5 May
2008–15 June 2008) were analysed. For these episodes, scientific focus, length, number
of sections and listener interactivity (how often hosts discussed listener feedback) were
recorded. Each topic covered was noted so that related/unrelated content could be identi-
fied during the analysis. If comments appeared to refer to episodes published before the
monitoring period, the relevant podcasts were located and checked to ensure that these
comments were indeed related.
Integrated online discussion forum (IODF) analysis
All comments posted during the iTunes monitoring period were analysed. The purpose
of the analysis was to determine what kinds of comments are being posted on blogs and
forums associated with science podcasts, and how these might stimulate interactivity and
discussion among posters. Content analysis is a well-tested approach to studying text-based
media (Stokes, 2003) and was therefore considered the most appropriate strategy within
this relatively new context. Full contents of each discussion facility were analysed,1
including comments made on discussion threads already running before the monitoring
period, as well as posts on newly initiated discussion threads.
Discussion on the IODFs was divided into five broad categories: podcast-related,
content-related, IODF-related, other relevant issues and irrelevant issues. Content-
related discussion refers to comments about specific topics covered during podcasts, as
opposed to podcast-related discussion, which refers to comments about a podcast in gen-
eral, including positive or negative comments about its production, suggestions for items
and technical difficulties. IODF-related content covers comments about the blog or
forum itself. Other relevant issues are issues within the relevant field of science but
which are not episode-specific. Irrelevant issues encompass spam and completely off-
topic conversation outside the scope of the podcast.
To assess whether users were truly sharing knowledge and opinions, or simply post-
ing individual feedback, an additional, smaller coding frame was developed to analyse
interactivity via relationships between comments. Under this coding frame, comments
that referred specifically to previous comments were quantified, as were instances of
naming and quoting other posters.
All content was coded using the program NVivo, which facilitated quantitative as
well as qualitative analysis. Commentary was coded thematically (see Mason,
2002). NVivo was used to calculate the number of comments within each category,
as well as a percentage of total content based on word count. Where appropriate,
qualitative examples were identified and extracted from relevant categories and
subcategories.
Recruitment for listener interviews
A number of strategies can be used to recruit interview participants online, such as: direct
contact, by locating email addresses; posting to user groups or on social network sites;
Birch and Weitkamp 895
recruiting through advertisements (see for example, Mann and Stuart, 2000: 26–27).
Each method has pitfalls; the researcher might be accused of sending spam or the mes-
sage might not be posted to a list serve if the moderator (a gatekeeper) feels it is not
appropriate. With this in mind, producers or hosts for each of the selected podcasts were
contacted and asked to advertise for participants on the researcher’s behalf. Text adverts
were provided that could be placed online to target listeners who also used forums.
A short screening questionnaire was used to identify respondents representing a range of
listening habits, both in terms of which podcasts they listened to and how often they
listened. Although 20 completed questionnaires were received, only three of the selected
podcasts were represented, with a strong bias towards Guardian Science Weekly.
Therefore, additional potential interviewees were individually targeted in order to fill
gaps through relevant listener groups on the social network site ‘Facebook’. Targets were
randomly selected by the first letter of their first names.
Finally, the entire sample was stratified by listenership, as well as frequency of con-
tribution to the forums. As none of the respondents contributed more than ‘occasionally’,
the sample was divided into two groups: contributing and non-contributing. Relationship
to the scientific community and gender were also considered when sampling from each
of the sub-populations. Because of the Science Weekly bias, some interviewees were
chosen specifically to increase the representation of respondents listening to other sci-
ence podcasts. However, during interviews it became apparent that in total seven of the
ten listened to science podcasts not being studied. The opportunity was therefore taken
to ask questions in the wider context of science podcasts.
Listener interviews
Neuendorf (2002: 52) asserts that it is inappropriate to make conclusions about sources
or receivers – in this case, contributors or listeners – based purely on a content analysis.
In line with this view, interviews were designed to fill gaps in knowledge that would be
left by the content analysis, including identifying characteristics of listeners who did not
contribute to forums. Questions focused on the practicalities of listening to podcasts and
reasons for contributing or not contributing to forums, as well as offline conversations
relating to the podcasts.
All interviews were transcribed fully and coded in NVivo by interview question and
interviewee. Common themes were identified. For instance, all answers for the question
‘Where are you generally listening?’ were compared and, as a result, answers referring to
travelling/commuting were grouped together.
Results
The podcasts
Table 1 shows the range of podcast formats covered. Shows vary considerably in length,
from just 11 minutes to around an hour, as well as in number of sections. In addition, not
all are published weekly. The Naked Scientists and WNYC Radio Lab podcasts are also
aired as radio shows on local public radio stations. Hosts did not mention IODFs more
896 new media & society 12(6)
than twice, on average, per show. Listener feedback (from all sources) tended to be incor-
porated to a similar extent, except on the Naked Scientists podcast, where the level was
much higher due to regular question and answer sections. However, the majority of this
feedback was contributed via email rather than the forum. It should be noted that
‘Directions to IODF’ refer specifically to a podcast’s IODF as opposed to its publisher’s
main website, although in some cases these were the same.
The Science Weekly and Naked Scientists podcasts tend to have more complicated, and
rigid, structures with a larger number of different segments. For instance, Science Weekly
generally begins with a rundown of what will be in the show and then proceeds to a news
section and often a pre-recorded interview interspersed with in-studio discussion among
hosts and guests. Topic changes are indicated, as they are on radio shows, with ‘sweepers’
(pre-recorded dividers – musical or voiced). In contrast, The Math Factor has no formal
intro/outro, or sweepers, and usually covers one specific mathematical problem and little
else. Astronomy Cast, the other more specialised podcast, also usually focuses on one main
topic, such as the Hubble Space Telescope. The other podcasts (all listed as ‘general sci-
ence’ podcasts in Table 1) cover a broad range of topics from the link between breast
feeding and intelligence (The Naked Scientists) to space tourism (Science Weekly). Radio
Lab episodes in the sample strayed outside the bounds of science into music and culture.
IODF analysis
In the sample studied, there was a large difference between the highest and lowest num-
bers of posters (see Table 2). Frequency of posting was generally quite low, with only
around a third (35.2%) on average posting more than once during the entire six-week
period. Interestingly, the proportion of posters who posted more than once is very similar
for the forums and substantially higher than for the blogs. The Radio Lab blog was
almost entirely populated by single post users, with only 10% posting more than once.
Table 1. General characteristics of selected podcasts
Podcast Science Shows in IODF type Length Sections Direct to Discuss
focus sample (min) IODF feed-back
Astronomy astronomy 6 forum & 31 2.17 1.67 0.17
Cast blog
Naked general 6 forum 58 16.83 1.00 9.50
Scientists science
WNYC’s general 3 blog 18 1.00 0.00 0.00
Radio Lab science
Science general 6 blog 38 7.83 1.67 1.67
Weekly science
The Math maths 5 blog 11 1.60 0.40 0.40
Factor
‘Sections’ are defined as topic changes. ‘Direct to IODF’ and ‘Discuss feedback’ are frequencies for hosts
directing listeners to their podcasts’ IODFs and for discussing listener feedback, respectively. All values are
means, except ‘Shows in sample’, which is a total for the period studied.
Birch and Weitkamp 897
The average length of posts ranged between 67 and 129 words. Post length was not asso-
ciated with numbers of posts, posters or one-time only posters.
Number of posts per poster followed a similar distribution across most podcasts
(see Figure 1). As shown, few posters posted more than ten comments during the moni-
toring period. It is worth noting that in the >11 category for the Astronomy Cast forum
there were individuals who posted 21, 39 and 69 times.
0%
10%
20%
30%
40%
50%
60%
>1 >2 >3 >4 >5 >6 >7 >8 >9 >10 >11
Number of posts
Posters
Astronomy Cast forum Astronomy Cast blog Naked Scientists
WNYC’s Radio Lab Science Weekly The Math Factor
Figure 1. Number of posts posted by individuals (posters) on each IODF
Table 2. Summary of IODF activity
Podcast Total % posting more Total Average Posts by
posters than once posts post length producers (%)
(words)
Astronomy Cast (forum) 59 52.5 260 128 3.4
Astronomy Cast (blog) 23 26.1 37 99 2.7
Naked Scientists (forum) 36 52.8 128 121 31.6
WNYC’s Radio Lab (blog) 39 10.3 41 67 2.4
Guardian Science Weekly (blog) 22 36.4 39 81 10.3
The Math Factor (blog) 12 33.3 18 129 16.7
Mean 32 35.2 87 104 11.2
Post and poster data includes comments made by producers. In the case of the Naked Scientists IODF,
which is part of a more general ‘Science Forum’, data were only collected from radio and podcast feedback
sections1 (this includes ‘Question of the Week’ posts and posts filed under ‘Radio Show & Podcast
Feedback’.)
898 new media & society 12(6)
The first category on the x axis shows the percentage who posted more than once, the
second shows the percentage who posted more than twice and so on. Producers are
omitted here.
Nature of discussions
Figure 2 shows how the nature of discussions differed between IODFs; the level of com-
pletely irrelevant discussion was very low in general, making up less than 0.5% of the
content in each case. The wider scope of discussion on forums than blogs was probably
due to the larger numbers of posters using the forums, as opposed to some factor unique
to forums. Also, there were only two forums and both were very active.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Astronomy Cast forum
Astronomy Cast blog
Naked Scientists
WNYC’s Radio Lab
Science Weekly
The Math Factor
IODF
Share of discussion
Irrelevant issues
Other relevant issues
IODF-related
Content-related
Podcast-related
Figure 2. Nature of discussions on IODFs
The largest portion of discussion fell into the content-related category. However,
posts on the Radio Lab blog were mostly related to the podcast itself and were also the
least conducive to debate. More than half of all Radio Lab comments were simple positive
(39.2%) or negative (11.3%) feedback comments.
Within the content-related category, much of the commentary fell under the ‘providing
further information’ sub-category, making up between 15.0 per cent and 38.4 per cent of the
total content on the Science Weekly, Astronomy Cast and Naked Scientists IODFs. Posters
often mixed further information with personal opinions about a particular scientific topic.
Birch and Weitkamp 899
Heard a question on astronomy cast that got me thinking … for the question: ‘is the universe
expanding?’ shouldn’t the only logical/observable/provable answer also be ‘in relation to
what?’ We’re comfortable saying the universe is expanding (though that is only in relation to
us; after all, the residents of the universe) … Isn’t it just as possible that all within the universe
are simply shrinking within the singularity?
The shrinking contents of a singularity, after a space-time ‘grid’ is formed within, might also
lead to an easier understanding of the ‘expansion’ of space without a center since people won’t
be imagining a ‘big-bang’ throwing things out from a point.
Poster A, Astronomy Cast forum
These kinds of comments are far more conducive to debate and interaction than those
concerning the podcast itself, and in this way, several of the IODFs seemed to be provid-
ing users with places to build knowledge and collaboratively work through ideas related
to those covered on the podcasts. On the Astronomy Cast forum, users frequently
swapped theories and discussed topics in more detail than in the podcasts themselves,
suggesting many users were highly scientifically qualified.
Comments in the content-related category on the Math Factor blog (65.8% of the
total) differed considerably in that they focused almost exclusively on working through
solutions to mathematical problems; posters used the forum to facilitate collaborative
problem solving. Indeed, most users appeared to be maths specialists and had often writ-
ten their own programs to carry out calculations.
I whipped up some Python code that any of you can hack into to try out the Collatz function.
Poster B, The Math Factor blog
For the robot puzzle, if we label the commands thus … then the following program will suffice …
Poster C, The Math Factor blog
The reason that average post length was particularly long on the Math Factor blog was
that posters sometimes copied long stretches of programming code and instructions into
these posts.
The problem of producers
It should be noted here that producers’ comments are included in the summary data. As
shown in Table 2 the visibility of producers on forums ranged from around 1 in every
3 to 1 in every 50 posts. On The Naked Scientists forum, for example, where producer
involvement is high, the number of posts made by members of the public is significantly
lower than the total. Also, because it is easy for posters to hide their identity online, it is
possible that some producers were masquerading as members of the public and were
therefore not identified during the analysis. These are important points because they
900 new media & society 12(6)
suggest that levels of discussion cannot be taken at face value, here or on any podcast
IODF where producers are active.
Contributions from producers mostly fell into the podcast-related category and
focused on technical aspects of downloading podcasts, including answering queries
about missing or malfunctioning content. However, on the Naked Scientists forum, pro-
ducers did sometimes become involved in debate about elements of content, as in the
following brief exchange between one listener and the host.
The explanation for pre-menstrual cravings made no sense whatsoever: at 28 days the body is
getting ready to flush out the unfertilized egg, not preparing for pregnancy (unless the egg is
fertilized, which is a whole other story.) My humble opinion is that cravings are culturally, not
physically induced.
Poster D, The Naked Scientists forum
On the contrary, I would argue that in order to maximise the chances of pregnancy the body
needs to ensure that it is vitamin and micronutrient replete BEFORE ovulation occurs so that
there is a highly receptive endometrium and healthy biochemical milieu to greet the released
egg and support migrating sperm. This will increase the likelihood of successful fertilisation,
migration and implantation (pregnancy).
Producer, The Naked Scientists forum
Interactions between posters
Several different variables were used to measure interactivity. However, it became clear
during the research that certain measures, such as how often posters quoted each other,
were influenced by the layout and customs peculiar to specific IODFs. Quoting other
posters was actively promoted on the Astronomy Cast forum via a ‘Quote’ button, so it
is perhaps unsurprising that the rate of quoting was high here compared to blogs where
no such facility exists. The most appropriate measure, therefore, was regarded as the
‘response rate’ to other posts, which although more subjective was more broadly appli-
cable. Response rate is defined here as how often posters post comments in response to
others’ comments. Astronomy Cast’s response rate was the highest at 0.84, meaning that
84 in every 100 posts were responding to previous comments by other posters. As shown
in Figure 3, interactivity increased with total numbers of posters.
The Radio Lab blog’s particularly low level of interactivity, and deviation from this
trend, may be due to the nature of the discussion on the blog, which focused on com-
ments about the podcast itself rather than on discussion of content.
Response rate also appeared to be associated with post length. When interactivity was
low, so was average post length; at high levels of interactivity, posts were far longer. The
Math Factor is a notable exception to this trend and may be explained by the lengthy
programming codes and instructions that increased word counts on this IODF. (If these
are omitted, average post length for the Math Factor blog drops to 77 words, bringing it
neatly in line with the trend).
Birch and Weitkamp 901
Although it is difficult to establish a causal relationship between interactivity and post
length based on such a small sample, it is easy to imagine how the two could be linked.
Post length, given the very low level of irrelevant discussion, may be considered a reli-
able indicator of depth or level of discussion and greater depth of discussion might be
achieved when posters are interacting more frequently. Alternatively, interactions
between posters might be stimulated by more in-depth discussions.
Listener interviews
Listener profiles. All interviewees had at least undergraduate degrees in science or a
closely related subject. As shown in Table 3, the sample was evenly split into two groups
in terms of contributions to IODFs – half occasionally contributed and half never did.
Three of the final ten were Australasian, five were British, one was American and one
had joint Trinidadian and British citizenship. More than half listened to other science
podcasts, but only three listened to more than two science podcasts in total. Nine of the
ten interviewees regularly listened to a number of different podcasts, and subject areas
were wide-ranging from news and sport to comedy and literature.
How people listen. Nine of the ten interviewees said they listened while doing other
things, most commonly commuting or walking, but also working or cleaning.
However, this flexibility seemed to be causing disconnects between listening and
contributing to blogs and forums. Two listeners recognised the fact that not being at
their computers while listening meant they did not think about contributing to forums
or blogs.
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
0 10 20 30 40 50 60 70
Total posters on IODF
Rate of response (per post)
WNYC
Radio Lab
The Math
Factor
Science
Weekly
Astronomy
Cast blog
Naked
Scientists
Astronomy
Cast forum
Figure 3. Interactivity increases with total numbers of posters
902 new media & society 12(6)
Table 3. Science listening profile of interviewees
ID Job class IODF Sex Astronomy Science Naked WNYC’s The Math Science Science Others
contributor Cast Weekly Scientists Radio Lab Factor Friday Talk
1 SC yes ♀ ••• ••• ••• ••• –
13 SC yes ♀ ••• –
17 SC no ♀ •• 4
18 UC yes ♂ ••• •• 3
19 UC yes ♂ ••• •• –
24 UC no ♂ ••• –
25 SC no ♀ •• –
27 UC no ♂ ••• ••• –
28 SC no ♂ • •• 1
29 SC yes ♂ •• –
SC = working in a science-related career.
UC = career unrelated to science.
♀ = female. ♂ = male. • = occasional listener, •• = regular listener, ••• = rarely misses an episode.
Birch and Weitkamp 903
I’m never thinking about it – because I’m always listening on the go, it just never occurs to me.
I rarely look at these site’s web pages.
Interviewee ID28
As well as spatial disconnects, one listener also mentioned the time lag between publica-
tion and listening as a reason not to contribute.
Online conversations. Although most listeners were aware that the blogs and forums
existed, none were regular contributors. It transpired that among the Science Weekly
listeners, several had misinterpreted the terms ‘blogs and forums’ as referring to the
discussion facility on Science Weekly’s Facebook page. However, as podcast producers
administrate this group, comments relating to these discussions are also of interest, par-
ticularly as some interviewees seemed to regard it as an easy way to access discussion
related to the show; this suggests there is a convenience aspect to social networking sites
that might make them valuable to producers in terms of accessing listeners and gaining
feedback. As podcasts are often downloaded via podcatchers, it might be more difficult
to drive listeners to a publisher’s website, which they would not otherwise need to visit.
Those who did contribute tended to do so only when a particular topic caught their
attention, for example a topic in which they already had an interest or about which they
held strong views. Others commented to correct inaccuracies in the programme or on the
IODF itself. Except for those making corrections, no listeners indicated they ever went
to a blog or forum with the specific intention of making a comment. In fact, listeners
often suggested they simply had no real inclination to contribute. No listeners were con-
cerned by the public nature of forums. Only one indicated that she was put off by the
level of scientific detail on IODFs and was the only interviewee who had a degree in a
closely related subject, as opposed to in a natural science.
These findings suggest that to encourage people to contribute more often, producers
might need to provide more creative incentives to drive people to their IODFs or devote
more time to highlighting particular discussions within the podcasts themselves.
I have actually subscribed to the RSS feed for Radio Lab’s comments, so I do actually get a sort of
stream of comments and I sort of glance at them and mostly just delete them. Some of them are quite
interesting, I mean at the moment, for example, Radio Lab is actually asking its listeners to recommend
their favourite episodes … So I’m sort of thinking, you know, what would I like to suggest?
Interviewee ID4
Offline conversations. All listeners said they discussed issues raised on science podcasts
with friends, colleagues or family members, although they were more likely to choose
people whom they knew were interested in science.
I talk to my father quite a lot … And he’s a scientist as well, at least by training, so … if I find
something interesting we’ll talk about it.
Interviewee ID29
904 new media & society 12(6)
Two interviewees said they would prefer to have conversations in real life than online,
which could limit the potential for interaction on IODFs.
Value of science podcasts. All interviewees said that science podcasts were valuable
sources of information about science. This is perhaps not surprising considering most
were regular listeners. However, the level of enthusiasm for science podcasts and the
medium in general was evident among several listeners, suggesting they felt they had
benefited immensely from using these tools.
I think it’s fantastic. I’m a big podcast fan and especially when it comes to science … If I
was to have to actively go and look for articles, or to look for things I’m interested in
I probably wouldn’t do it. It’s an easily digestible, well-presented way of getting quality
information.
Interviewee ID24
Although enthusiasm for the forums was far less evident, in general, interviewees from
both groups – those who had previously contributed and those who hadn’t – thought that
the forums were an important addition to science podcasts. Interviewees viewed the
IODF not only as an opportunity to provide feedback to producers but also to contribute
to debate and to the podcast itself.
The Guardian Science, they do really interact well with the blogs and I’ve contributed a few
times on the blogs and it’s been picked up in the podcast and they do work quite hard to
integrate what’s happening on the forums … and I think that interaction makes it much more
real and it just feels like the debate is happening in a sense and I like that.
Interviewee ID18
This shows listeners like to know they have opportunities to discuss what they hear on
podcasts, or to make comments about the way they are produced, even if they don’t neces-
sarily use them. As most listeners said they read comments but did not always contribute
to discussions themselves, there is also a level at which blogs and forums may be useful for
stimulating internal thought processes about science and possibly offline conversations.
Discussion
There is a general move towards strategies that increase debate and deliberation about new
scientific advances (European Commission, 2006; House of Lords, 2000). The findings
from this study suggest that science podcasts could play a role in creating and informing
such debate and discussion. However, for science-focused podcasts such as those in this
study, these conversations are likely to be instigated only between those with a prior inter-
est. Listening figures were not available for all podcasts, and in some cases were confiden-
tial; however, sources suggest audiences for at least two of the podcasts reach into the tens
Birch and Weitkamp 905
of thousands per episode (Gay, 2007; Green, 2008). If this is the case, numbers of IODF
contributors in this study certainly represent less than 1 per cent of audiences over a six-
week period, clearly indicating barriers to contribution. These could include a reluctance
to participate, as found by Thurman (2008). Several listeners in our study indicated that
they read comments written on IODFs but do not contribute themselves.
Our research suggests that one reason for the low number of posters is a spatiotem-
poral disconnect created by the flexibility and portability of the podcast medium itself.
Portability is clearly a benefit for listeners in terms of access to podcasts; however, it acts
as a barrier to contributing to online discussion forums. Listeners must access the inter-
net to contribute to IODFs and are not usually positioned to do so while listening to the
podcast itself; by the time of their next internet visit they will have forgotten or lost the
inclination. In this sense, some of the listener-control aspects of podcasts touted as
advantages could actually act as disadvantages in terms of creating discussion about sci-
ence topics and this may limit the potential of podcasts to stimulate meaningful discus-
sion on their associated online discussion forums. This contrasts with the findings of
Schmitz-Justen and Wilhelm (2007), who found that discussion forums could be devel-
oped which would promote knowledge exchange among interested users.
The problem of access to the internet when listening on portable media devices may
be remedied during the next few years by the increasing quality of mobile internet ser-
vices and media convergence. Listeners will start to download podcasts to their mobile
phones, many of which already function as MP3 players and internet browsers. In the
meantime, podcast producers could get around the spatiotemporal disconnect by provid-
ing more obvious incentives to visit IODFs (‘look at this cool picture on our blog’), by
highlighting contributions from individual users (‘Joe Bloggs from Texas said …’) and
by providing a wider variety of channels for feedback, such as via forums on popular
social networking sites, which may be considered more ‘convenient’ than publishers’
own websites. The present study suggests that interactions between individuals increase
as the community using a discussion facility grows in size, so by making concerted
efforts to drive up traffic on their blogs and forums, producers can increase information
sharing. By creating a larger community, a kind of ‘Wikipedia effect’ should come into
play – theoretically, more contributors means there should be less chance of erroneous
information passing unnoticed, as argued by Jenkins (2006).
The vast majority of content on blogs and forums associated with the science podcasts
studied was considered to be highly relevant. This goes some way towards dispelling
Minol et al.’s (2007) concerns that online discussion forums are filled with inappropriate
content, at least where science podcasts are concerned. Most comments related specifi-
cally to topics covered in podcasts, although on more highly populated forums, users
were more likely to discuss related topics. This suggests IODFs could be important in
broadening interest among listeners and suggests the emergence of the ‘intelligent’ com-
munities for information exchange envisaged by Lévy (1997). In this study, knowledge
sharing was an important function of some IODFs, particularly the Math Factor blog,
where programming code was often posted for broader use. These developments indicate
that interactive discussion forums could be considered ‘local communities’ based on
shared web space and subject interest, as suggested by Postill (2008).
906 new media & society 12(6)
The visibility of hosts and producers on IODFs varied widely between podcasts, but
they were active in each case. Their presence in these online environments and inclusion
of listener contributions in podcasts shows that these tools are able to facilitate commu-
nication between the media and the public in a way that has not, traditionally, been pos-
sible. The interactivity of podcasts should not be overstated; radio phone-ins have for a
long time provided opportunities for media consumers to feed into the media itself.
However, IODFs attached to podcasts do offer advantages over live phone-ins, perhaps
the most important being that they give users more time to consider their actions and
more control over the subjects discussed.
On most IODFs, hosts and producers only intervened in conversations to answer tech-
nical queries. However, their heightened presence on The Naked Scientists forum raises
an interesting issue about expertise. Arguably, media producers will be more knowledgeable
than their listeners about the subjects featured on their podcasts – mainly because they will
have had to undertake research before recording and, possibly, spoken directly to experts
involved in the fields covered. (In some cases, hosts may also be scientists themselves.)
Their contributions can be considered extremely valuable, but their very expertise may
reinforce existing hierarchies between scientific experts, the media and listeners. The
extent to which producers should be encouraged to participate in discussions on ‘public’
forums could be questioned and the effects that such intervention has on listeners would
be an interesting avenue to explore further. For example, do producers augment the debate
or does continual media interference diminish the sense of public ownership of the space?
As McDermott (2008) and Schmitz-Justen and Wilhelm (2007) suggest, the interac-
tive functionality of Web 2.0 provides opportunities to share information. The blogs and
forums associated with podcasts offer opportunities to explore scientific ideas in greater
depth and, crucially, converse with other listeners. In this study, it is shown that interac-
tions often occur more frequently on blogs and forums with higher numbers of posters,
but on some, the most active people appear to be either scientists or otherwise highly
qualified in science. This does not devalue podcasts or IODFs as tools for science com-
munication – it is just as important to create opportunities for the scientifically literate to
discuss scientific ideas as it is for ‘lay’ people. In addition, it increases the level of scien-
tific knowledge that can be accessed within an online community, such that lay people
may learn from ‘experts’. One concern, however, is that those who do not have scientific
backgrounds may feel excluded from such communities and this issue warrants further
investigation. Barriers such as access may reduce contributions to online forums, but this
is not the only way that podcasts could stimulate discussion and there was evidence from
the interviewees that the podcasts did spark offline discussion. This is an important point
for two reasons: firstly, it means that podcasts could be useful in stimulating offline con-
versations about science; secondly, it shows that studies of online behaviour cannot rely
solely on publicly available information.
This study has combined data about podcasts with listener interviews and content
analyses of blogs and forums to interrogate the value of podcasts in stimulating discus-
sion about science. This triangulation approach offers a robust research methodology in
a new and challenging area. However, the small sample sizes studied meant that it was
difficult to identify trends in activity on blogs and forums or among listeners. The study
Birch and Weitkamp 907
also identified variables that could not be analysed within the scope of this project. For
example, posters on the forums (as opposed to the blogs) generally displayed their loca-
tion alongside their name and it was apparent that users often interacted with people from
other states, countries or continents. However, because this information was not avail-
able on each IODF, it was not included in the analysis. In further work, it might be inter-
esting to fix selection criteria for podcasts to include availability of geographical
information and use this as a starting point for research into cross-cultural influences on
shared knowledge.
Conclusions and recommendations
Online discussion forums associated with science podcasts provide listeners with an
opportunity to explore scientific issues further. As such, they could contribute to wider
political agendas related to increasing public deliberation and discussion about scientific
advances, although it is clear from this study that only a small minority of listeners contrib-
ute to forums. These tools also offer opportunities for audiences to help shape the media
agenda, although at present their use is limited. This pilot study serves as a basis on which
future studies of podcasting as an innovative means of science communication could build.
The following recommendations are made for further studies:
• Larger, better controlled studies of conversations taking place on discussion facil-
ities associated with podcasts, for instance comparing same-sized communities
and podcasts covering similar subject areas.
• Detailed tracking of the activities of individuals, where possible taking into
account geographical location and scientific expertise of posters, as well as of those
they interact with.
• Research exploring the roles of ‘experts’ within online communities, including
media producers and scientists.
• More in-depth analyses of the scientific content of IODFs, involving relevant
experts.
• Studies involving interviews with the most active contributors.
• More wide-ranging surveys of media consumption by podcast listeners.
Acknowledgements
We would like to acknowledge the support and collaboration of Robert Frederick, Dr Pamela L.
Gay, Ben Green and Ben Valsler. We would also like to thank the anonymous reviewers for their
helpful comments.
Note
1 ‘Full contents’ refers to all comments rather than a random sample. In some cases IODFs were
part of larger forums not specifically related to podcasts and therefore only podcast-related
sections were analysed.
908 new media & society 12(6)
References
Anderson C (2004) The Long Tail. Wired, 12.10, October 2004, URL (consulted Feb. 2008): http://
web.archive.org/web/20041127085645/http://www.wired.com/wired/archive/12.10/tail.html
Beaudoin CE (2008) The internet’s impact on international knowledge. New Media & Society
10(3): 455–474.
Berry R (2006) Will the iPod kill the radio star? Convergence 12(2): 143–162.
DeLong D (2007) The Journalist Article Donnacha DeLong 25 October 2007, URL (consulted Feb
2008): http://donnachadelong.blogspot.com/2007/10/journalist-article.html
Eggington B (2008) Podcasting and New Media. Stempra, London, 25 February 2008.
European Commission (2006) Final Report of the Study of the Integration of Science and Society
Issues in the Sixth Framework Programme. European Commission, EUR 22976, 1–112.
Frederick R (2008) Personal communication, 28 May 2008.
Friess S (2006) Axing the Podcast Middleman. Wired, 10 March, URL (consulted Feb 2008):
http://www.wired.com/science/discoveries/news/2006/10/71892
Gay PL (2007) Astronomy Cast: Evaluation of a podcast audience’s content needs and listening
habits. Communicating Astronomy with the Public, 1: 24–29.
Gay P, Bemrose-Fetter R, Bracey G, and Cain F (2008) Astronomy Cast: Evaluation of a podcast
audience’s content needs and listening habits. Communicating Astronomy to the Public 1(1): 24–29.
Green B (2008) Personal communication, 27 May 2008.
Green MC, Brock TC, and Kaufman GF (2004) Understanding media enjoyment: The role of
transportation into narrative worlds. Communication Theory 14(4): 311–327.
Griffiths P (2008) Is Digital Radio the New BetaMax? Reuters, 12 February 2008, URL (consulted
July 2008): http://blogs.reuters.com/uknews/2008/02/12/is-digital-radio-the-new-betamax/
Holtz S (2005) Rationalizing a Business Podcast. Global PR Blog Week, URL (consulted Oct 2008):
http://www.globalprblogweek.com/2005/09/20/holtz-rationalizing-business-podcasting/
House of Lords (2000) Science and Society. London: Her Majesty’s Stationary Office.
Jenkins J (2006) Convergence Culture: Where Old and New Media Collide. New York: New York
University Press.
Keen A (2007) The Cult of the Amateur: How Today’s Internet is Killing Our Culture. London:
Nicholas Brealey Publishing Ltd.
Kiss J (2008) Internet Radio Hits the Mainstream. Guardian.co.uk, 28 January 2008, URL (con-
sulted Feb. 2008): http://www.guardian.co.uk/media/2008/jan/28/web20.digitalmedia
Knowledge Agency, The (2004) The iPod Generation. OFCOM/The Knowledge Agency, 23 July
2004, 1–49.
Leigh P (2005) Future Developments in Podcasting. Inside Digital Media, Inc. URL (consulted
Sept. 2009) http://insidedigitalmedia.com/research.php
Lévy P (1997) Collective Intelligence: Mankind’s Emerging World in Cyberspace. Cambridge
MA: Perseus Books.
McDermott N (2008) Podcasting and New Media. Stempra, London, 25 February 2008.
McLaughlin L (2006) Podcasting 101: What the web’s new trend means to you. IEEE Pervasive
Computing 5(4): 4–6.
McLuhan M (1962) The Gutenberg Galaxy. Toronto: University of Toronto Press.
Mann C, Stuart F (2000) Internet Communication and Qualitative Research: A Handbook for
Researching Online. London: Sage.
Birch and Weitkamp 909
Mason J (2002) Qualitative Researching, 2nd edn. London: Sage.
Merzagora M (2004) Science on air: the role of radio in science communication. Journal of Science
Communication 3(4): 1–6.
Minol K, Spelsberg G, Shulte E, and Morris N (2007) Portals blogs and co.: The role of the internet
as a medium of science communication. Biotechnology Journal 2: 1129–1140.
Neuendorf K.A (2002) The Content Analysis Guidebook. Thousand Oaks, CA: Sage.
Newholm T (2008) The digital divide and optimal slack. New Media & Society 10(2): 295–296.
Picardi I and Regina S (2008) Science via podcast. Journal of Science Communication 7(2):1–4.
Postill J (2008) Localizing the internet beyond communities and networks. New Media & Society
10(3): 413–431.
RAB/IAB (2006) Using Radio with Online Radio Advertising Bureau in association with the
Internet Advertising Bureau UK, Multi-media Effectiveness: 4, 1–23.
RAJAR/Ipsos MORI (2008) Podcasting and Radio Listening via Internet Survey. RAJAR, January
2008, 1–20.
Schmitz-Justen FJ, Wilhelm AFX (2007) An empirical study of factors impacting on knowledge
processes in online forums: Structural equation modelling analysis and results. International
Journal of Web-based Communities 3(3): 252–270.
Stokes J (2003) How to do Media and Cultural Studies. London: Sage.
Thurman N (2008) Forums for citizen journalists? Adoption of user generated content initiatives
by online news media. New Media and Society 10(1): 139–157.
Vorderer P, Klimmt C, and Ritterfeld U (2004) Enjoyment: At the heart of media entertainment.
Communication Theory 14(4): 388–408.
Hayley Birch is a freelance science communicator. She writes for and edits magazines,
websites and popular science books, produces podcasts with Sounds of Science and is
involved in EU science, society and policy initiatives. She received her MSc in Science
Communication from the University of the West of England.
Emma Weitkamp explores new media approaches to communicating science, both from
research and practitioner perspectives. Projects include the use of new media to com-
municate environmental science, as well as creating an online science comic for primary
schools. She received a PhD in Biochemistry from Cambridge University.
