Tracking the Impact of Media on Voter Choice in Real Time: A Bayesian Dynamic Joint Model

TL;DR: A Bayesian zero-inflated dynamic multinomial choice model is developed that enables the joint modeling of: the interplay and dynamics associated with the individual voter's choice intentions over time, actual vote, and the heterogeneity in the exposure to marketing communications over time.

Abstract: Commonly used methods of evaluating the impact of marketing communications during political elections struggle to account for respondents’ exposures to these communications due to the problems associated with recall bias. In addition, they completely fail to account for the impact of mediated or earned communications, such as newspaper articles or television news, that are typically not within the control of the advertising party, nor are they effectively able to monitor consumers’ perceptual responses over time. This study based on a new data collection technique using cell-phone text messaging (called real-time experience tracking or RET) offers the potential to address these weaknesses. We propose an RET-based model of the impact of communications and apply it to a unique choice situation: voting behavior during the 2010 UK general election, which was dominated by three political parties. We develop a Bayesian zero-inflated dynamic multinomial choice model that enables the joint modeling of: th...

Summary

1 Introduction

• Recent advances in information technology, along with the advent of social media and enablement of user-generated content, have resulted in a proliferation of consumer data.
• Importantly, the authors distinguish between exposure to communications that are controlled and paid for by a firm or party (‘paid’ media such as advertising, newspaper inserts, and billboards) and those that are not controlled by them (‘earned’ media such as news, editorials, communications about a particular brand by rival brands, etc.).
• The authors explore the potential of a real-time experience tracking (RET) technique in addressing many of the aforementioned weaknesses.
• The respondents, who receive prior training about sending these brief, very structured messages, provide information on three essential aspects of their encounter: the brand name, the communications type (for example, TV advertisement or press editorial), and their perceptual response to the encounter.

2 Motivating Data

• Many organizations continue to rely on cross-sectional surveys for collecting data, particularly in politics.
• One approach used by market research agencies such as Ipsos MORI is self-reported influence, which is based on asking respondents what media influenced them in their voting decisions (Worcester et al. 2011).
• A second approach is to correlate recalled exposure or perceptual response to different media with brand attitudes (O’Cass 2002).
• The use of a panel responding to repeated surveys, such as the American National Election Studies database, arising from interviews conducted before and after each presidential election (Klein and Ahluwalia 2005), addresses the issue of common method bias but not the problem of recall bias.
• It is well known that the time of voting decision is important and has been highlighted as a possible segmentation variable, not least because those who decide how to vote during a campaign are affected by media campaigns (Fournier et al. 2004).

2.1 Real-time Experience Tracking (RET)

• Figure 1 illustrates how RET works by showing the data collection process for the politics study.
• The pre-survey also captures brand intention at the start of the study.

2.2 The Campaign: The 2010 UK General Election

• The above mentioned RET method was used during the 2010 UK general election to collect their motivating data.
• The initial survey captured the respondent’s voting history: here, whether they had voted in the last general election, and if so, what their vote had been.
• Thus, each respondent’s was provided online at the end of the pre-survey and then the coding scheme was summarised by text message to the respondent for easy access when texting about exposures.
• The reason for this is that broadcast political advertising is illegal in the UK and ubiquitous in the US.

3 A Zero-Inflated Joint Model of Media Communications and

• The authors joint model consists of five components that capture the impact of media exposures on choice.
• 2) a zero-inflated multinomial logit model to capture the final vote choice given the decision to vote.
• Zero-inflated distribution is used in order to account for the excess zeros caused by the fact that respondents are typically not exposed to the majority of marketing communications on a daily or regular basis.
• 5) finally, the authors join the models using a correlated random effects approach.
• Another motivation for joint modeling is the fact that even when the authors are interested in understanding how choice intentions depend on exposures, they need to jointly model since these exposures are not exogenously determined.

3.1 Dynamic Multinomial Logit Model for Choice Intentions

• To build the model for the intention to vote for a particular political party during any week of the campaign, the authors note that voters were faced with four primary choices, viz., ‘Labour’, ‘Conservatives’, ‘Liberal Democrats’ or ‘Do Not Intend to Vote’.
• The authors denote the random voter’s heterogeneity effect for the ith voter and kth party as bik.
• The `1 and `2 are coefficients of interaction between earned media and paid media for frequency and valence respectively.
• The authors formulate a dynamic model with the lagged term (yi,t−1) incorporating the presence of any stickiness in voters’ choices from previous periods.
• There are several reasons why these coefficients might vary over time, e.g., due to changes in exposures, changes in the policy of the party, or occurrence of unexpected events which may lead the voters to think differently.

3.2 Zero-Inflated Multinomial Logit Model for Final Vote

• Since the final choice (vote) took place at the same time for all respondents, the data for the final vote is not time varying.
• As explained earlier, the authors model the decision of whether or not to vote with a binary distribution and the conditional choice decision as a multinomial logit model, resulting in a zero-inflated multinomial model.
• The authors also include the factor of whether the voter intended to vote for any major parties in previous weeks.
• The, at is the effect of the tth week’s voting intention on the final vote and since these effects will also have diminishing effects over time, the authors assume that at = a × φ3−t.
• Here a is the overall effect and φ measures the decay over weeks.

3.3 Model for the Frequency and Valence of Communications Exposures

• The authors include both the frequency and the average valence of communications exposures within this model to separately capture the effect of the number of exposures and the encounter valence in the model.
• Second, any associations that are influenced by the message may grow in attitude strength through familiarity (Erdem and Keane 1996).
• The valence of communications may also have an impact on the choice of the party.
• Since the presence of these excessive zeros leads to spurious over-dispersion (Park et al. 2011), the authors fit a zero-inflated Poisson (ZIP) regression model which is a discrete mixture of true zeros (degenerate distribution) and positive counts for frequency of exposures (Lambert 1992).
• Note that although exposure valence is measured on a five-point scale, in their data the authors have these scores averaged over each week, which makes xVitkl a continuous variable.

3.4 Correlation Structure and Heterogeneity: Combining the Models

• All the three models described above carry information about the voting pattern of the voters and are therefore inter-related.
• Since these outcomes are measured on a variety of different scales (viz., multinomial, ZIP), it is not possible to directly model the joint predictors’ effects due to the lack of any natural multivariate distribution for characterizing such a dependency.
• Moreover, without inter-relating or jointly considering these outcomes, it is hard to answer questions about how the evolution of one response (e.g., intention to vote) is related to the evolution of another (e.g., final vote).
• A flexible solution is to model the association between different responses by correlating the random heterogeneous effects from each of the responses.
• Due to the IIA property of the multinomial logistic model, it is not possible to find cross effects it is not possible to incorporate cross effects (e.g., how a negative exposure to Labour may affect the voting intention for the Liberal Democrats) directly in the utility equation (3).

4.2 Prior Specification and Posterior Inference

• The authors estimate the model parameters using a Bayesian framework.
• The authors now describe the prior distributions for the model parameters.
• A similar approach has been used in mixed effects model settings (Natarajan and McCulloch 1998).
• The sampler from the posterior obtained from the Markov Chain Monte Carlo (MCMC) simulation allowed us to achieve a summary measure of the parameter estimates.

4.3 Model Selection and Model Fit

• Before discussing their results, the authors first compare their proposed model with some alternative models to test the quality of model fit.
• The authors also compute Log-Pseudo Marginal Likelihood (LPML) as an additional model selection criterion and Posterior Predictive P-value for model fit.
• The summary statistics from the predicted and observed data are given by χ2(y, θg) and χ2(yrep,g, θg), respectively, where yrep,g denote the replicated value of y from the posterior predictive distribution of θ at the gth iteration of the Gibbs sampler.
• From the above results, the authors see that for their proposed model (model-2) the DIC4 value is the lowest and the LPML value is the highest.

5.1 Results for the The Impact of Paid and Earned Media Exposures on

• Voting Intention Table 2 depicts the results of the lagged variables on the voting intentions model.
• The authors find that the valence of both paid and earned media has a positive and significant impact on voting intentions (with the exception of earned media valence for week 2 and paid media valence for week 1, which are non-significant).
• This effect is more prominent with less well known brands, where attitudes are not already accessible (Goh et al. 2011).
• Further work is required to check these observations about valence wearout, which ideally would be studied over longer periods and in alternative decision contexts.
• Parameter estimates for the impact of demographics and socio-political issues on voting intentions also provide some interesting insight.

5.1.1 Impact of Increased Exposure Frequency and Valence

• Having established the importance of these media effects, the authors assess the impact of marginal increases in exposure frequency or valence on choice behavior.
• Z are other control variables and b are random effects.
• The authors can compare effects of changes in exposure frequencies using βF1,t and β F 2,t and exposure frequencies for other media as well as compare the values of βF·,t for different values of t.
• For illustrative purposes, the authors show the impact of a small increase in earned and paid media frequency and valence for the average voter with 30%, 50% and 80% probability of voting for party 3 (Liberal Democrats) at one, two and three weeks prior to the final vote.
• The authors find that the marginal effect of improved earned media valence is positive and, by some margin, highest in week 1.

5.2 Results from the Final Vote model

• Table 6 contains the results from the logistic regression analysis of whether or not a person voted.
• The positive and significant coefficient φ2 suggests that those who voted in 2005 were also more likely to vote in the current election, as were those who intended to vote for one of the three major parties (φ3).
• As expected, the decay parameter tells us that the most recent decisions have the greatest impact.
• As exposure valence continues to have an effect on intentions in each week, this emphasizes the need to prioritize the quality of media exposures right up until the final choice decision.
• Also, a being positive and significant reaffirms their assumption that the effectiveness of these exposures depreciates over time.

5.3 Results from the Exposure Frequency and Average Valence Model

• The results from estimating exposure frequency and valence (Tables 9 and 10) provide several interesting results.
• Those from lower social categories (manual, semi-manual and lowest grade workers) have higher paid media frequencies (ζF11,2); this may result in part from their greater television consumption.
• Older age voters report higher frequencies particularly of earned media (ζF13,1), and they report a lower valence for these earned media exposures (ζV13,1).
• People in employment report higher exposure frequencies (ζF16,1, ζ F1 6,2).
• This corresponds with their higher likelihood of voting, particularly for those in the public sector (Corey and Garand, 2002).

5.4 Correlation Matrix

• The estimated value of the correlation matrix is given in Table 11.
• This is the estimate of the variancecovariance matrix from section 3.4 and equation 12.
• These cut-offs reflect somewhat arbitrary limits, and results should be considered in the context from which they are derived.
• The authors find a very weak positive linear association between exposure valence of paid media and exposure frequency of earned media (corr (hVi1, h F i2) = 0.168, significant) indicating that how positive or negative a paid communication is perceived to be is only weakly related with how often earned media is experienced.

5.5 Simulation Study

• The simulation is focused on evaluating the finite sample performance of the proposed Bayesian estimation when the data are generated, mimicking the real data the authors have.
• Based on the results, the authors find that the estimates under the proposed model reliably recovered the true estimates with reasonable coverage probability.

6 Conclusions

• The authors have reviewed the relative strengths and limitations of existing methods of evaluating voters’ response to communications, proposed the use of RET for this purpose, presented a dynamic model of the impact of communications on voters’ choice, and applied the model to a three-way political party choice.
• This once again emphasizes the managerial importance of understanding communications valence and not just exposure, contrary to the accepted gold standard in commercial practice of market mix modeling based on exposure or proxies for it, and contrary to the common managerial emphasis on ‘share of voice’ metrics.
• While this has many benefits as compared with the cross-sectional survey, clearly a brand manager would ideally want insight into drivers of sales, not just attitude shift: after all, their final choice model shows that it is not just the final week’s attitudes which shape the actual decision.
• Overall, the authors believe that RET could have a profound impact on the measurement of communications effectiveness, given its ability to model individual consumer response to communication encounters and, importantly, evaluate the relative influence of those encounters.

Figures

Table 6: Decision to Vote: Final Vote Model

Table 1: Summary Statistics 1

Table 2: Parameter Estimates-Dynamic Coefficients (Intention to Vote Model).

Table 5: Sensitivity Analysis for Average Level of Encounters

Figure 1: RET Data collection.

Table 4: Average Frequency and Average Valence of Encounters

Table 10: Average Valence Model : Demographics

Table 9: Model Exposure Frequency : Demographics

Figure 3: Reporting of encounter from each party each week.

Table 8: Parameter Estimates of Party Exposures (Final Vote Model)

Figure 2: Reporting of paid and earned encounters each week. Earned encounters are encounters with the brand (party) which are not controlled by the political party itself, such as news and editorial comment. More people are exposed to earned communications.

Table 7: Previous Intentions (Final Vote Model)

Table 3: Parameter Estimates for Exposures (Intention to Vote Model)

Table 11: Posterior estimates of correlation matrix Σ̂.

Full text

Tracking the Impact of Media on Voter Choice in Real
Time: A Bayesian Dynamic Joint Model
Bhuvanesh Pareek, Pulak Ghosh,
Hugh N. Wilson, Emma K. Macdonald, and Paul Baines
September 9, 2017
Bhuvanesh Pareek is Assistant Professor, Department of Operations Management and Quantitative Techniques
at Indian Institute of Management, Indore, India; Pulak Ghosh is Professor, Department of Decision Sciences and
Information Systems, Indian Institute of Management, Bangalore, India; Hugh N. Wilson, Emma K. Macdonald and
Paul Baines are Professors of Marketing at Cranfield School of Management, Cranfield University, UK; Big thanks to
Prof. Rajiv Sinha who has been very instrumental in this work, although he is no more with us, his contribution will
be cherished.
1
Journal of the American Statistical Association, Available online December 15 January 2018
DOI:10.1080/01621459.2017.1419134
Published by Taylor & Francis. This is the Author Accepted Manuscript issued with: Creative Commons Attribution Non-Commercial License (CC:BY:NC 4.0).
The final published version (version of record) is available online at DOI:10.1080/01621459.2017.1419134 Please refer to any applicable publisher terms of use.

Abstract
Commonly used methods of evaluating the impact of marketing communications during
political elections struggle to account for respondents’ exposures to these communications due
to the problems associated with recall bias. In addition, they completely fail to account for the
impact of mediated or earned communications such as newspaper articles or television news,
that are typically not within the control of the advertising party, nor are they effectively able to
monitor consumers’ perceptual responses over time. This study based on a new data collection
technique using cell-phone text messaging (called real-time experience tracking or RET) offers
the potential to address these weaknesses. We propose an RET-based model of the impact of
communications and apply it to a unique choice situation: voting behavior during the 2010 UK
general election, which was dominated by three political parties. We develop a Bayesian zero-
inflated dynamic multinomial choice model that enables the joint modeling of: the inter-play
and dynamics associated with the individual voter’s choice intentions over time, actual vote,
and the heterogeneity in the exposure to marketing communications over time. Results reveal
the differential impact over time of paid and earned media, demonstrate a synergy between the
two, and show the particular importance of exposure valence and not just frequency, contrary
to the predominant practitioner emphasis on share-of-voice metrics. Results also suggest that
while earned media have a reducing impact on voting intentions as the final choice approaches,
their valence continues to influence the final vote: a difference between drivers of intentions
and behavior that implies that exposure valence remains critically important close to the final
brand choice.
Key Words: Advertising; Bayesian Methods; Communications - paid and earned; real-time
experience tracking; Zero-inflation; Multinomial Logit Model; Voters’ Choice;
2

1 Introduction
Recent advances in information technology, along with the advent of social media and enablement
of user-generated content, have resulted in a proliferation of consumer data. Additionally, the wi-
despread use of mobile devices by consumers makes it feasible to collect data about influences on
consumer choice in real time. This presents an opportunity for researchers wanting to model the
effects of communication on consumer choice, such as organizations working to better understand
the voter’s choice during an election campaign. However both the volume of consumer data being
generated and its real-time nature necessitates the use of new and original analytical tools. This
creates fresh opportunities for informed managerial decision-making, particularly for firms wor-
king to better understand the voter’s choice during an election campaign. This paper addresses
two substantive issues concerning the impact of marketing communications on voting choice: the
differential effectiveness of alternative marketing communications, and the relative importance of
communications frequency on the one hand and the valence of the consumer’s perceptual response
on the other. Global spend on advertising in 2017 was $583.91bn(eMarketer 2017). Communicati-
ons evaluation is a substantial business activity in its own right, providing around 15% of the trade
for the $68bn market research industry(ESOMAR 2016). Many firms also work hard to generate
positive communications that are mediated by third parties such as journalists and retailers. Much
of this effort aims to influence the consumer’s choice of brand, whether it be consumer products
and services, business-to-business suppliers, non-profit organizations, or political candidates and
parties; the latter being the context for this study. Given the massive expense involved and their
potential impact on consumer choice, evaluation and optimization of the effectiveness of commu-
nications is critical. While there have been several studies analysing election data (Linzer 2013,
Christensen and Florence 2008, Gelman et al. 2008), to our knowledge no study has examined
the relative influence of multiple paid and earned media communications on a voter’s choice of a
political party during an election period.
This study makes use of a new and innovative data collection procedure that permits individu-
als to report on communications from competing political parties as soon as they encounter them,
thereby alleviating the recall bias associated with traditional research techniques. Importantly,
we distinguish between exposure to communications that are controlled and paid for by a firm or
party (‘paid’ media such as advertising, newspaper inserts, and billboards) and those that are not
controlled by them (‘earned’ media such as news, editorials, communications about a particular
brand by rival brands, etc.). These are referred to as paid and earned communications respectively.
Although these are important modes of communication, finding the impact of these communicati-
3

ons on voters’ choice has four key challenges. The first is the typical problem with survey based
data. Although surveys can explore voters’ attitudes towards a political party and can often, but
not always, be quite helpful in forecasting election results (Linzer 2013), most respondents find
it difficult to recall their exposure to communications accurately. One can examine the impact of
media spend on party choice, but only rarely are data available on the impact of earned communi-
cations such as media coverage (which is not controlled by political parties), and it is often difficult
to disentangle the effects of the medium from the message. The second problem is in assessing
the differential impact of various communications that are controlled by the brands. Models based
on media spend are extensively used by media planners to examine the relative impact of different
paid media such as TV, radio, and billboards. Although the impact of some earned media such
as newspaper editorials has been similarly assessed, this stream of work generally uses entirely
separate data sources and models (Goh et al. 2011, Stephen and Galak 2012). However, marketing
directors and political strategists need to allocate resources between paid media on the one hand
and PR and service improvements focused on influencing earned media, on the other. Third, in
assessing communications impact, researchers frequently ignore the crucial role of the consumer’s
perceptual response to the communication. Theoretical and experimental work confirms that atti-
tude towards communications influences attitude towards the brand (Bri
˜
nol et al. 2004), but most
models of field data, particularly in the case of paid media, focus purely on communications ex-
posure (Verhoef 2003) or the proxy of media spend (Naik and Peters 2009), presumably because
perceptual response data are frequently unavailable. Fourth, practitioners have an interest in un-
derstanding how the impact of communications on voters’ intentions evolve over time and prior to
the final brand/party choice. In many consumer choice contexts, particularly in a voting scenario,
consumers have extended purchase journeys, within which the timing of communications may be
important. Furthermore, communications such as advertising are often intended to have medium-
term impacts via attitude change as well as more immediate behavioral impacts; similarly, PR may
deliberately or accidentally have lasting effects. Traditionally, a key area of interest has been how
to model shifts in voting intention or preference over time and how these shifts impact on subse-
quent choice decisions. However, what has not been researched, up to now, is why these preference
changes have occurred and what role paid and earned communications play in that attitude change
process.
In this article, we explore the potential of a real-time experience tracking (RET) technique in
addressing many of the aforementioned weaknesses. RET has recently been adopted for commu-
nications evaluation by a number of multinationals such as Unilever, BSkyB, PepsiCo, HP, Ener-
gizer, InterContinental Hotels and 20th Century Fox, among others (see Macdonald et al. 2012).
4

The genesis of this technique may be traced to the need for capturing the most essential aspects
of consumer encounters with brand communications in real time and in a quick and easy response
format. This research method involves asking a panel of trained respondents to send a cell phone
text (SMS) message as soon as they encounter any of a set of competitive brands during the study.
The respondents, who receive prior training about sending these brief, very structured messages,
provide information on three essential aspects of their encounter: the brand name, the commu-
nications type (for example, TV advertisement or press editorial), and their perceptual response
to the encounter. Respondents also complete a brand attitude survey at the start and end of the
study. RET can capture the impact of a wide range of communications on an individual in real
time; however, no model of brand choice is currently available to exploit the novelty of these kind
of data. We use the RET method to collect data on choice of political parties in the UK’s 2010
General Election, with a tracking period - the length of time for which each participant is asked
to report their communications exposure-of the complete four-week election period. In addition to
reporting all party (brand) encounters throughout the month by text message, we asked participants
to provide their initial voting intentions, and to provide weekly updates on their voting intentions,
via weekly surveys. For modeling purposes, this context has three benefits: the final party choice
is made simultaneously by all participants, thereby alleviating problems of data censoring; the
2010 UK election campaign spanned a four-week period from the announcement of the election to
the actual vote, allowing us to track intentions for the entire choice cycle for all participants; and
it enables the capture of paid and earned communications on the choice of political party for an
individual in a way that many existing research techniques cannot.
Consequently, we propose a RET-based model to capture the impact of these unique, real-time
encounters on the choice of political party. Our model incorporates both paid and earned commu-
nications, and takes into account both the frequency of exposures and their valence, a measure of
the voter’s perceptual response to the communication. The ability to track, differentiate and model
these two effects is one of the most important contributions of this research. We present a Bay-
esian Zero-Inflated dynamic multinomial choice model for simultaneously analyzing four critical
factors that are associated with the impact of communications on voters’ choice: (a) voter’s choice
intentions and their evolution over time (b) voters’ final choice (c) the frequency and valence of
paid and earned media, and (d) heterogeneity in the exposures to different communications over
time.
In summary, this article contributes to the statistical literature on voter choice theory and, more
broadly, on consumer choice in four ways: (1) it unpacks the differential impact of paid and earned
communications during a particular purchase window, thereby allowing firms to optimize their re-
5

Frequently asked questions

Q1. What have the authors contributed in "Tracking the impact of media on voter choice in real time: a bayesian dynamic joint model" ?

In addition, they completely fail to account for the impact of mediated or earned communications such as newspaper articles or television news, that are typically not within the control of the advertising party, nor are they effectively able to monitor consumers ’ perceptual responses over time. This study based on a new data collection technique using cell-phone text messaging ( called real-time experience tracking or RET ) offers the potential to address these weaknesses. The authors propose an RET-based model of the impact of communications and apply it to a unique choice situation: voting behavior during the 2010 UK general election, which was dominated by three political parties. Results also suggest that while earned media have a reducing impact on voting intentions as the final choice approaches, their valence continues to influence the final vote: a difference between drivers of intentions and behavior that implies that exposure valence remains critically important close to the final brand choice. 

Q2. What are the future works mentioned in the paper "Tracking the impact of media on voter choice in real time: a bayesian dynamic joint model" ?

RET studies would allow practitioners to check for this possibility. To extend such a survey to a multiple-week RET study up until the purchase would enable the model the authors have described to be applied with very few modifications. To support practitioners in the appropriate application of RET, further research is needed to better understand its validity properties, and to locate its best role in the market researcher ’ s armoury. Further work is also needed to fuse RET with other data sources such as objective purchase recording, using retailer loyalty card data for example.