A leading question is one which leads or persuades a person towards giving a particular response. The way in which a question is ‘framed’ - the use of particular wording and limiting the range of acceptable answers - can direct a person to giving an answer that lacks the accuracy of one provided in response to an open-ended question.
For example, the following question may be considered to be leading:
Did you see the man in the black-and-white overalls?
The use of the definite article in the question (“the man” instead of “a man”) implies that there was a man in black and white overalls. The respondent may be inclined to positively confirm their sighting of such a man, as they may assume that the man was definitely present, but they may have missed him by their own lack of observance. In reality, however, the man may not have been on the scene or existed at all.
The use of leading questions has wide-ranging implications in numerous aspects of life, and raises doubts as to the reliability of some court convictions where an eyewitness testimony may have been influenced by biased interviewing techniques.
The framing of questions can also be crucial to referendum results, in which voters may be inclined to vote for a particular outcome depending on the wording of the question printed on their ballot paper.
Research in the early 1970s, led by cognitive psychologist Elizabeth Loftus, brought attention to the reliability of eyewitness testimonies, demonstrating that an interviewer could influence interviewees’ responses by manipulating the phrasing of a question. In an oft-cited 1974 experiment, she found that witnesses to a car crash could be persuaded to testify to false memories, such as seeing a car travel at different speeds, depending on the wording used to describe the resulting crash. Assumptions regarding the severity of the crash, suggested in question wording, act as a kind of retrospective interference on witnesses’ memories of an event (Loftus and Palmer, 1974).
Types of Leading Question
Leading questions can solicit a particular response using a number of methods. Phrasing may be adjusted to suggest a particular answer, or the interviewer may limit the range of permissible responses, sometimes excluding the natural answer of the respondent. Examples include:
Did you steal Jack’s cake?
Were you at the club?
Leading vs Loaded Questions
Whilst a leading question leads a person to give a particular answer, a loaded question (also known as a complex question) contains the assumption that one or more additional assertions are true. For instance, the assumption that a defendant was at a crime scene may be made by reference to it in a loaded question:
Why did you steal the cake from the shop?
Here, the act of cake theft is assumed not questioned - but the motive for it is.
At what time last night did you drive away in the getaway car?
The interviewer refers to the person driving away in the getaway car in their request for a time.
What were the two of you arguing about that ended in you hitting him?
The interviewee is implied to have hit a victim without him/her having confessed to it.
In their answer, the respondent is expected to share the assumed fact with the interviewer without questioning it.
A tag question may be used to suggest a fact or opinion, and then ask that the respondent confirms it:
You were at the club on Friday, weren’t you?
The fireworks are great, aren’t they?
Whilst open-ended questions (e.g. “Where were you last night”) invite any response, closed-ended questions lead the respondent to a limited choice of answers.
A dichotomous question offers one of two responses, e.g. “yes or no”, “true or false”, “good or bad”, omitting more nuanced answers.
Multiple choice questions provide a wider range of answers for a respondent to choose from, but can still lead them to a particular set of responses:
What is your favorite flavor of ice cream: strawberry, vanilla, neapolitan or chocolate?
The question above would prevent lovers of toffee ice cream from reporting their preference accurately. Authors of questionnaires which use multiple choice questions, such as those involved in market research, must take care not to influence or lead participants in such a way that their responses could be skewed.
Referendum Question Framing
The framing of questions can be a significant issue for those assessing referenda. Prior to the 2015 Scottish independence referendum, the Scottish government (in favor of independence) proposed that the following question be presented to voters:
“Do you agree that Scotland should be an independent country? Yes/No” (Electoral Commission, 2013)
The Electoral Commission objected to the phrasing of the question, suggesting that “the words ‘Do you agree’ potentially encouraged people to vote ‘yes’ and should be replaced by more neutral wording”. Ultimately, the leading question was replaced with:
“Should Scotland be an independent country? Yes/No” (Electoral Commission, 2013)
Leading upto the 2016 referendum on European Union membership, the Electoral Commission again took the view that the proposed question could lead respondents to give a particular answer.
A yes/no question was proposed: “Should the United Kingdom remain a member of the European Union?” However, this was again contested as favoring a confirmatory ‘remain’ vote, and the government accepted that the question should be replaced with, “Should the United Kingdom remain a member of the European Union or leave the European Union?”. Voters were given two options - to “Remain a member of the European Union” or to “Leave the European Union” (Electoral Commission, 2015).
In the 1980s, amidst fears that police questioning techniques were producing inaccurate or unreliable eyewitness testimonies, an alternative set of methods were proposed for use during interviews, with the intention that the undue influence of leading questions on witnesses’ recollection of events could be reduced.
In a 1985 paper, R. Edward Geiselman suggested that witnesses be encouraged to relive a previous event, recalling unrelated facts regarding a crime scene, such as the weather and any passers-by, so that they may more accurately experience and recall memories specific to a particular crime.
Along with further suggestions, this was termed the ‘cognitive interview’, and was found to produce more accurate recollections of events than standard interview methods. The suggestions made in the cognitive interview have had a lasting impact on modern police techniques used during eyewitness interviews (Geiselman, Fisher et al, 1985).
Intentionally or accidentally, leading questions can impact on the testimonies provided by eyewitnesses in trials, influence referendum outcomes and affect the accuracy of survey results. Beyond the phrasing of questions, a number of other factors can also affect the answers given to questions.
The environment in which questions are asked can weigh on respondents’ behavior. In a series of group conformity experiments, psychologist Solomon Asch asked a group people to respond individually to an obvious question. When confederates in the group provided an incorrect answer, genuine participants tended to follow them, for fear of deviating from the social norm and being disproven (Asch, 1955). Asch demonstrated that we often take into account the responses of our peers when considering our own answers to questions, even when we feel that fellow respondents are mistaken.
By building a rapport with a respondent, interviewers can also solicit a desired response to a specific question.
Using a method known as the foot-in-the-door technique, if an interview can persuade a person to respond positively to a series of questions, they are more likely to be able to solicit a ‘yes’ answer to a further question. The technique was tested in an experiment by Jonathan Freedman and Scott Fraser of Stanford University in 1966, and may explain why so-called ‘cold-callers’ who make unsolicited sales calls tend to ask friendly questions that a person would be inclined to agree with, prior to inviting them to sign up for a service or order a product (Freedman and Fraser, 1966).