Core conflict in associations: meaning and application

Your association data show two strong associations around the same choice that point in opposite directions, such as freedom and the hassle of charging for electric car sharing. What does such a clash mean for the choice and for your communication? This article explains what Neurofactor means by a core conflict in associations and how to recognise and test it.

A core conflict in associations is a Neurofactor working concept for two strong, salient associations around the same choice that clash, usually a clearly positive and a clearly negative one.

It is not an established scientific term but a working definition and a hypothesis that you test with interviews and a follow-up measurement. Research on approach-avoidance conflict, attitudinal ambivalence and cognitive dissonance provides context, but does not validate the working concept.

An association is a link between a subject, such as a brand or service, and a meaning, feeling or image; see Association: meaning and application. Associations differ in strength, salience and valence: how positively or negatively charged they are.

Neurofactor working definition: a core conflict in associations is a situation in which, within a defined choice context, two associations around the same choice are both strong and salient but opposite in valence, so they can pull the choice in different directions.

• Strength. Both associations are firmly linked to the subject; see Association strength.
• Salience. Both stand out and come to mind quickly in the choice context; see Association salience.
• Opposite valence. One association is clearly positive, the other clearly negative.

'Core' means that both associations are central to how people see the choice. The term comes from Neurofactor's research practice, not from the literature. We therefore treat a core conflict as a hypothesis, not as an established cause of behaviour.

Four lines of research provide context; none has studied the Neurofactor concept itself.

Approach-avoidance conflict

Distinguishing types of conflict, such as attraction versus aversion around the same goal, is usually traced back to Kurt Lewin (Lewin, 1935). Neal Miller summarised experiments measuring how strongly animals pulled towards a reward or away from a shock at different distances. As a negative event came closer in time or space, aversion increased more steeply than attraction to a positive event (Miller, 1944, summarised by Rozin & Royzman, 2001).

Attitudinal ambivalence

Holding positive and negative evaluations of the same subject simultaneously is called attitudinal ambivalence. Thompson, Zanna and Griffin proposed an index in which ambivalence increases as positive and negative reactions become stronger and more similar in size (Thompson, Zanna & Griffin, 1995). Priester and Petty distinguished such an 'objective' approach from subjectively felt inner conflict. According to their model, felt ambivalence increases with the number of reactions on the smaller side, first quickly and then more slowly, and decreases as the dominant side grows larger (Priester & Petty, 1996).

Van Harreveld and colleagues found in three studies that ambivalence is unpleasant only when someone has to commit to one side; uncertainty about the outcome of the decision explained the link with physiological arousal (Van Harreveld et al., 2009a). In a review, they connect ambivalence with decision-making and different ways of coping (Van Harreveld, Van der Pligt & De Liver, 2009b).

Cognitive dissonance, briefly

Leon Festinger described cognitive dissonance as tension between beliefs, or between a belief and behaviour, that cannot be reconciled (Festinger, 1957). Dissonance research often concerns what happens after a choice; a core conflict describes tension in the associations around the choice.

Negativity bias, with nuance

In a broad review, Baumeister and colleagues conclude that bad events, information and impressions often carry more weight than comparable good ones. They refer to research by Kahneman and Tversky in which losing money caused more distress than gaining the same amount caused pleasure, and they also write that many good experiences can outweigh a single bad one (Baumeister et al., 2001). Rozin and Royzman call the negativity bias 'far from universal'; positive biases exist as well (Rozin & Royzman, 2001).

Negative associations may therefore carry extra weight in a core conflict. The literature does not support the claim that the brain automatically chooses the least risky option. What people do must be studied per choice and context.

Look for a pattern in a defined group's data within one choice context:

1. Two associations with high strength, mentioned by many, not by a handful of people.
2. Both with high salience: they come up early or spontaneously.
3. Opposite valence: one clearly positive, the other clearly negative.
4. Within the same group and the same choice context.
5. In the same people: the strongest signal is that many participants mention both associations.

If one half mainly mentions the positive and the other half mainly the negative association, you probably have a divided group: two patterns side by side, possibly two associative target groups. That calls for segmentation, not for resolving one conflict. Because intermediate forms also occur, always look at the level of individuals. On clusters, see What are association clusters?.

With a deal-breaking association, one negative association wins and the option is dropped. With a core conflict, the positive association also remains strong and the choice stays open.

1. Define the choice context: which choice, for whom, at what moment.
2. Describe each association separately: strength, salience and valence, with the measurement method; see How do you measure associations?.
3. Check the individual level: how many participants mention both associations, and do they form a recognisable subgroup?
4. Conduct in-depth interviews: what do people mean by both associations, and does the tension really play a role in the choice?
5. Run a follow-up measurement: does the pattern recur in a new sample? Where possible, add a measure of felt inner conflict and a choice or intention measure.
6. Record what you know: confirmed, weakened or rejected, for which group and under which conditions.

Even a confirmed core conflict remains a description of associations, not proof of a causal effect on behaviour.

The obvious response is to promote the positive association more strongly. With a core conflict, that often falls short, because the negative association remains just as strong and salient. According to ambivalence research, the tension is also felt mainly when choosing.

So start by acknowledging the conflict. Name the negative side as people experience it and show what is being done about it or what to realistically expect. This only works with something concrete to offer. A message that talks away a real problem can increase distrust.

These are editorial recommendations, not researched effects of this working concept. Test messages against outcome measures defined in advance.

This example is fictional and has no research results.

Situation. A provider of electric shared cars wants city residents to give up their own car. A first association measurement among people considering this shows two prominent ones: 'freedom' and 'hassle with charging'.

Choice question. Is this a core conflict that slows down switching, or are these two different groups?

Available information. Both associations are mentioned often and early, with opposite valence. How many mention both is unknown.

Suitable approach. The team analyses the answers per participant. If many of the same people mention both, there may be a core conflict. If mainly different people each mention one side, it looks more like a divided group. Interviews clarify what 'hassle' means, such as finding a charging point or uncertainty about range. A follow-up measurement tests whether the pattern recurs and relates to the intention to switch.

Possible interpretation. If it is a core conflict, a campaign that only promises 'freedom' is probably not enough. A message that honestly addresses charging and explains how charging concerns are reduced is a hypothesis to test.

Next step. Compare two message variants on measures set in advance, such as understanding, perceived concern and intention.

• Calling every mixed picture a core conflict. A weak or barely salient negative association does not make a core conflict.
• Using group totals as evidence. Without analysis per person, you cannot distinguish a core conflict from a divided group.
• Presenting the working concept as a proven mechanism. Theories of ambivalence and conflict do not validate this working model.
• Predicting a fixed outcome. People facing such a conflict do not always postpone, drop out or choose the least risk.
• Translating animal experiments directly. The approach-avoidance gradients largely come from animal research.

A core conflict in associations is a Neurofactor working concept for two strong, salient associations with opposite valence around the same choice. Research on conflict and ambivalence makes it plausible that such tensions matter, especially when choosing; what people do varies. Treat a suspected core conflict as a hypothesis, check it per person, test it with interviews and a follow-up measurement, and take both sides seriously in communication.