When guilt works: a comprehensive meta-analysis of guilt appeals

Introduction Guilt appeals are widely used as a persuasive approach in various areas of practice. However, the strength and direction of the persuasive effects of guilt appeals are mixed, which could be influenced by theoretical and methodological factors. Method The present study is a comprehensive meta-analysis of 26 studies using a random-effects model to assess the persuasive effects of guilt appeals. In total, 127 effect sizes from seven types of persuasive outcomes (i.e., guilt, attitude, behavior, behavioral intention, non-guilt emotions, motivation, and cognition) were calculated based on 7,512 participants. Results The analysis showed a small effect size of guilt appeals [g = 0.19, 95% CI (0.10, 0.28)]. The effect of guilt appeals was moderated by the theoretical factors related to appraisal and coping of guilt arousal, including attributed responsibility, controllability and stability of the causal factors, the proximity of perceiver-victim relationship, recommendation of reparative behaviors, and different outcome types. The effect was also associated with methods used in different studies. Discussion Overall, the findings demonstrated the persuasive effects of guilt appeals, but theoretical and methodological factors should be considered in the design and testing of guilt appeals. We also discussed the practical implications of the findings.

that does not belong to the categories above.D. Study location (1 = US; and 2 = Non-U.S.; and *3 = Unknown).

Operational definition and coding instructions:
This variable records the location where the study was conducted.If the sampling location is different from that of the affiliated university of the authors, code the sampling location.If no location is reported in the article, the location of the affiliated university of the first author is coded.

Operational definition and coding instructions:
This variable records the year when the study was published.

F. Sample size
Operational definition and coding instructions of variable E, F, G, and H: The coders record the information provided by each study that describes the percentages of women, non-White participants in the sample, mean age of the sample, and the sample size.

Guilt appeal variables include:
A. Responsibility Attribution (1 = Responsibility attributed in a guilt appeal; 2 = No responsibility attributed in a guilt appeal/description of harm or suffering only):

Operational definition and coding instructions:
The variable records whether the guilt appeal in a study attributes the suffering of others to the perceiver's action or belief.
(1) Responsibility attributed in a guilt appeal: The guilt appeal in a study attributes the suffering of the victims to the perceiver's action or belief.
(2) No responsibility attributed in a guilt appeal/Description of harm or suffering only: The guilt appeal in a study describes the harm on victims or their suffering but does not attribute the harm to the perceiver's actions or beliefs.This variable is to record whether the guilt appeal in a study attributes the victim's suffering to a cause that is temporary and changing.
(1) Unstable causes attributed in guilt appeals: The guilt appeal in a study attributes the victim's suffering to a cause that is temporary and changing.
(2) No unstable causes attributed in guilt appeals: The guilt appeal in a study does not attribute the victim's suffering to a cause that is temporary and changing.If the study does not explicitly indicate the stability attribution, this option is selected.This variable records the social distance between the perceiver and the victim as described in the study.

D. The Proximity of Perceiver-Victim
(1) Self/Personal relationship: The study describes the victim to be self or someone close to the perceiver, such as a family member or friend.
(2) Work/Professional relationship: The study describes the victim to be a workplace co-worker, supervisor, subordinate, or in any other professional relationship with the perceiver.
(3) Community/Acquaintance: The study describes the victim to be someone that the perceiver knows slightly but does not have a professional or personal relationship with.The victims may also be someone that lives in the same neighborhood with the perceiver.
(4) Strangers: The study describes the victim to be someone that the perceivers have not met or known before.The victim has no direct connection or relationship with the perceiver.
(5) Other/Non-human: The study describes the victim to be a non-human entity, such as the natural environment.The variable is to record whether the guilt appeal in a study recommends any counteracting behavior or measures intended to undo the perceiver's wrongdoing and reduce its harm on victims.
(1) Yes: The guilt appeal in a study recommends at least one counteracting behavior or measure to undo the perceiver's wrongdoing and reduce its harm on victims.
(2) No: The guilt appeal in a study does not recommend any counteracting behaviors or measures to undo the perceiver's wrongdoing and reduce its harm to victims.

Operational definition and coding instructions:
The variable records the specific type of outcome variables that are measured in a study.
(1) Guilt: The study measures the emotional feelings of guilt as an outcome.
(2) Knowledge: The study measures participants' levels of knowledge or their understanding of a particular object.
(3) Attitude: The study measures participants' opinions or feeling about a persuasion object.
(4) Behavioral intention: The study measures participants' readiness or willingness to perform a behavior.
(5) Actual Behavior: The study measures whether or not participants have actually performed a behavior.This variable records the circumstances or situation on which the subject, stimulus event, and purpose of guilt appeal are based.For example, a guilt appeal that advocates recycling behaviors is related an environmental context.
(1) Advertising/Marketing: The guilt appeal in a study is related to selling products or services.
(2) Education: The guilt appeal in a study is related to obtaining knowledge in general or teaching and learning in a school or non-school setting.
(3) Environment: The guilt appeal in a study is related to protecting the natural environment.
(4) Medical/Health-related: The guilt appeal in a study is related to medical services for physical and/or mental wellbeing at a hospital or non-hospital setting.
(5) Political-related: The guilt appeal in a study is related to the government, legislation, or the way a country/region is governed.
(6) Safety instruction: The appeal in a study is related to reducing the risk in daily lives or that of operating machinery.
(7) Other: The guilt appeal in a study is related to a context that is not mentioned above.

Operational definition and coding instructions:
This variable records the initial processes of eliciting guilt in guilt appeals.Specifically, coders check how the study matches the core appraisal pattern of guilt to arouse the emotion.
(1) Persuasive message: The study provides each participant with a brief piece of information that is produced on a specific medium to match the core appraisal pattern of guilt.
(2) Recall: The study asks each participant to remember an incident in their own lives that once aroused guilt and then to re-experience it again now.
(3) Other: The study uses an induction method that does not belong to the first two categories.

Operational definition and coding instructions:
This variable records whether the study arouses guilt by using a story versus by using reasoning, logical arguments, and statistical information.
(1) Narrative: The study provides stories involving fictional characters and settings to arouse guilt.If the study asks the participants to write or recount their own experiences through storytelling, this option is also coded.
(2) Non-narrative: The study uses reasoning, logical arguments, and statistical information to arouse guilt.If the study asks the participants to write logicl arguments by themselves, this option is also coded.

Operational definition and coding instructions:
This variable records the modality of the stimulus or induction materials used in a study as a guilt appeal.Modality refers to the type of output channel or media through which the guilt appeal information is delivered from the experimenter to the participants as perceivers.
(1) Text only: The study uses written words only in the stimulus material of a guilt appeal.
(2) Text and image: The study uses both written words and static visual elements in the stimulus material of a guilt appeal.
(3) Text and video: The study uses both written words and moving visual elements in the stimulus material of a guilt appeal.This variable records the time when the study measures outcome variables after presenting guilt appeals.The measurement time points may vary between immediate to several days after guilt appeals occur.
(1) Immediate: The study measures outcome variables immediately or within a day after guilt appeals occur.
(2) Within a week: The study measures outcome variables about 1-7 days after guilt appeals occur.
(3) After a week: The study measures outcome variables more than 7 days after guilt appeals occur.
Note.Categories marked with an asterisk (*) were excluded from data analysis due to zero or very few cases coded.standard deviation of the pre-tests for treatment and control groups; and are standard deviation of the post-tests for treatment and control groups (Morris, 2008).
If no comparison group was used (i.e., within-subject experiment design), the standard deviation mean (g) was computed as (Lipsey & Wilson, 2001), where and are pre-test and post-test means of outcome variable; is the pooled standard deviation of pre-test and post-test; is the correlation between pre-test and post-test score.In cases where the correlation between pre-test and post-test score was not reported, of 0.5 was used to compute g as a default (Lipsey & Wilson, 2001).
For post-test control group (i.e., between-subject) design, the standardized mean difference (g) was computed using the formula , where is the pooled standard deviation of and , which is computed as ,where and are standard deviation and and are outcome for treatment group and control group.In cases where mean and standard deviations of each group were not reported, test statistics such as F and t values were used to calculate Hedges' g (Rosenthal, 1994).

( 6 )
Emotion other than guilt: The study measures a specific emotional feeling, excluding guilt, as an outcome, such as joy.(7) Motivation: The study measures motivation as an intermediate process that directs emotional arousal and cognitive or behavioral activities in response to external stimuli.(8) Cognition: The study measures a specific cognitive process or outcome related to acquiring and understanding information and knowledge.(9) Efficacy: A person's perceived ability or power to generate an effect, such as efficacy of conducting cervical cancer screening.(10) Other: The study measures an outcome that does not belong to any of the categories above.G. Context (1 = Advertising/Marketing; 2 = Education; 3 = Environment; 4 = Medical/Health-related; *5 = Political-related; *6 = Safety instruction; and 7 = Other); Operational definition and coding instructions: J. Modalities (1 = Text only; 2 = Text and image; *3 = Text and video; *4 = Image only; *5 = Video only; 6 = Audio only; *7 = Text and audio; 8 = Unspecified).

( 4 )
Image only: The study uses static visual elements only in the stimulus material of a guilt appeal.(5)Video only: The study uses moving visual elements only in the stimulus material of a guilt appeal.(6)Audio only: The study uses sound only in the stimulus material of a guilt appeal.(7)Text and audio: The study uses written words and sound the stimulus material of a guilt appeal.(8)Unspecified: The study does not use a modality that is listed above.This option also applies to the study that does not provide details regarding the modality of the guilt appeal.K. Time Points of measurement (1= Immediately after/within a day; 2 = Moderate delay/1-7 days; 3 = Long delay/More than 7 days):Operational definition and coding instructions: If the study does not explicitly indicate the responsibility attribution, this option is selected.This variable records whether the guilt appeal in a study attributes the victim's suffering to a cause that the perceiver has control over.