Conflict Measurement and Assessment

Ways of Assessing Conflict

There are several factors that make assessing conflict a challenge. Although conflict is a normal component of almost any interpersonal relationship, most people do not experience conflict on a daily basis, and when conflict does occur, people often try to keep it hidden from public observation. Therefore, it is especially difficult for researchers to directly observe conflict in a natural environment. Instead, researchers most typically assess conflict by making observations in a structured laboratory setting or by administering questionnaires that ask people to provide a self-report description of their conflict experiences.

Research with married couples has made extensive use of structured laboratory settings to assess conflict, and this method can be easily adapted to assess conflict in a wide variety of relationships. In the typical protocol, a couple visits a research laboratory where they are asked to identify an area of conflict. Partners may be separated and individually asked to describe a recent conflict episode, or they may be given a list and asked to identify issues that cause conflict in their relationship, or a researcher may interview the couple together and identify a common area of conflict. Using one of these techniques, a single area of conflict is selected as a topic for discussion, and the couple is then left alone in a room with instructions to attempt to resolve the issue. Typically, the couple is given 10 to 15 minutes for their conversation, and sometimes, [Page 304]couples are asked to complete two different conversations, discussing separate topics selected by each partner. With the couple's knowledge and consent, their conversations are video-recorded.

The video-recoded conflicts are scored using one of several possible approaches. Typically, researchers use a coding manual that provides precise definitions for different types of communication behavior. Trained coders then use detailed instructions in the coding manual to assign scores to the conversation. In this approach, it is essential to ensure that ratings are reliable. That is, the scores assigned to any given conversation should not be dependent on who coded it. Reliability can be improved by having each conversation coded by several different coders and also by ensuring that coders receive sufficient training. It is not always necessary, however, to train coders. Recent research has found that untrained coders can actually make reliable ratings regarding the extent to which people express certain basic emotions, such as anger or sadness. Another way to obtain scores is to have a couple rate their own conversation. For example, participants may watch a playback of their conversation and turn a dial in real time to indicate their level of emotion, or they may be asked to identify points in the conversation when a particular type of thought, feeling, or behavior occurred.

...