Considering whether a participant in a planned study will be a "subject at risk" or a "subject at minimal risk," according to recognized standards, is of primary ethical concern to the investigator. (APA, 1982, p. 5)
The investigator must carefully consider all possible threats to the well-being of research participants. Examples of serious threats include: "invasion of privacy, breach of confidentiality, longer term stress and discomfort, loss of self-esteem, upsetting reactions to being deceived or debriefed, embarrassment or humiliation, negative effects from being in a no-treatment control group, and reactions to being induced to commit reprehensible acts" (Koocher & Keith-Spiegler, 1998, p. 425). Unfortunately, predicting occurrence of these events is no easy matter, given the infinite variety of ways people may respond to situations arranged by the investigator. For example, a dirty joke that may be "amusing" to one participant may be "disgusting" and possibly even psychologically damaging to another. The general theme of Principle B, risk potential, is carried forward and elaborated in Principle G and Principle I.
Suppose an investigator plans to administer unpredictable loud noises to participants while they learn a task. Assessing risk potential would involve contemplating factors such as: the likelihood that the apparatus will malfunction in some way, exposing participants to dangerously high levels; the extent to which the noise may be detrimental to persons with certain pre-existing conditions (e.g., susceptibility to panic attacks); and, the possibility of negative collateral responses both during and after the experiment (e.g., many studies have shown that noxious stimuli elicit aggression: (Azrin, 1967).
When deciding for or against conducting a study, the investigator considers the risks for participants relative to the potential benefits that may accrue from the research. In general, greater risks are tolerated the greater the anticipated benefits. Unfortunately, as Koocher & Keith-Spiegler (1998) note, the benefits may be "virtually impossible to estimate accurately" (p. 424). A research outcome is uncertain; that is why the experiment is conducted in the first place. They describe two other problems. First, a benefit "may be in the eye of the beholder" (p. 425). Suppose an extinction (ignoring) procedure is shown to be a very effective means of reducing children's nighttime waking and crying. Some may argue that this enhances the children's self-reliance; others may argue that this damages the parent-child bond. Second, who or what is the object of the benefit? Should the focus here be on the research participant, the investigator, the knowledge base, or all of these? An intervention that successfully alters participants' eating patterns may make them less susceptible to heart disease, result in scholarly publication for the investigator, and provide support for or against a certain theory of behavior change. How does one weight each of these prospective positive outcomes in relation to risk tolerance?
Sally is concerned about how the young children in her study could react to the stranger during the Generalization Probe. Might some children become scared and inconsolable? Worse yet, could there be a lasting negative emotional effect, such as a generalized fear of all strangers? As part of her risk assessment, Sally surveys the literature. She discovers two studies very similar to the one she is proposing, both of which reported the children's behavior under a no-treatment condition: in one, nearly all the children immediately agreed to go with the confederate; and, in the other, three-quarters agreed to go with him and the rest did not run away. This suggests to Sally that being approached and enticed by an adult stranger is extremely unlikely to upset the children. The study is conducted as planned. One child, in response to the enticement, runs away from the stranger (as she learned to do in the video), but in the wrong direction, away from the school and into a busy street about 20 feet away. Sally's graduate students had chosen to conduct the confrontation near the street, reasoning that having a car within the child's sight makes the enticement more believable. They made this decision on the spot, soon after arriving at the school to conduct the Generalization Probe. Their only direction in this regard from Sally was for the stranger to approach the child in the schoolyard. Neither they nor Sally had ever visited the schoolyard.
Sally is concerned about how the young children in her study could react to the stranger during the Generalization Probe. Might some children become scared and inconsolable? Worse yet, could there be a lasting negative emotional effect, such as a generalized fear of all strangers? As part of her risk assessment, Sally surveys the literature. She discovers two studies very similar to the one she is proposing, both of which reported the children's behavior under a no-treatment condition: in one, nearly all the children immediately agreed to go with the confederate; and, in the other, three-quarters agreed to go with h im and the rest did not run away. This suggests to Sally that being approached and enticed by an adult stranger is extremely unlikely to upset the children. A week before the study begins, Sally visits the schoolyard to survey it for natural hazards and choose a safe location. She decides against one area of the playground -- although it would be very convenient for observational purposes -- because running toward the school would require that the children run down a steep slope, at the bottom of which is a stone parking lot. The children could easily fall and hurt themselves. Sally's graduate students suggest that they conduct the confrontation near the street so that the confederate could point to the car when enticing the children. They argue that this might make the enticement more believable. Sally agrees with their argument, but upon further brainstorming, they realize that the children could be in danger from street traffic if they were to run in the wrong direction. Sally considers enlisting the help of another graduate student, who could be positioned on the sidewalk during the confrontation so as to easily intercept the children should they run out into the road. She decides against this option because it could involve physically restraining the children, a potential added source of stress for them. In the end, she chooses a location near the swings, where natural hazards seem to be minimized.
The first item is an example in which there is inadequate risk assessment. (Sally could also be faulted for poor training of her graduate students; see Principle C.) Consulting the literature about how the children would likely react is a good first step. However, Sally should have visited the school playground in advance of running the experiment to determine how the setting itself could be dangerous. The second item is an improvement because this is what she does. From a research design perspective, two locations have distinct advantages: better reliability of observation nearby the hill, and better face validity nearby the street. However, for Sally, neither advantage justifies the potential risk associated with it. Sally makes compromises in the design of her study in order to safeguard the children's well-being (see Principle G).
One way to further assess the risks would be for Sally and her graduate students to role play the confrontation at different locations on the school playground, attempting to anticipate all possible reactions by the children, no matter how remote.