全球智库信息集成服务系统

Audiences recognized and liked the CVE content...

• More than a third of participants reported that they recognized the CVE content from both hashtags, suggesting that the live campaign reached a sizeable audience.
• Audiences reported liking the tested CVE posts at levels comparable with the mainstream advertising and PSA placebo content.
• There were differences in audiences' interpretations of the hashtag campaigns. Audiences appeared to understand the meaning of the #AkuTenmanu campaign, but appeared more confused about the messages of #CapekGakSih.

But the effects are uncertain

• The assessment did find a significant effect between the treatment and control groups. There was a significant effect on a measure of "promoting inclusivity in daily life on social media," but the positive effect did not stem from any change in treatment group opinion but rather from a significant, abrupt drop in control group scores on this measure.
• There was also a significant and rather strong negative effect on feelings about living in separated (or segregated) communities and judging others' appearances.

This report presents the results of an evaluation designed to assess the effects of countering violent extremism (CVE)–themed social media content used in a campaign to promote tolerance, freedom of speech, and rejection of violence in Indonesia. RAND Corporation researchers studied the effects of the campaign by recruiting a sample of Indonesian youth on Facebook and randomly assigning them to a treatment condition that exposed participants to CVE social media posts or a control condition. This report details the research design and findings and offers recommendations for improving such evaluations in the future.

The group Search for Common Ground (SFCG) worked with a market research firm to design content for Facebook, Instagram, and Twitter using two different hashtags developed specifically for the CVE campaign: #AkuTemanmu ("I am your friend") and #CapekGakSih ("Aren't you tired?"). RAND researchers recruited 1,570 participants from Indonesia via a series of Facebook advertisements. They assigned participants either to a treatment group that viewed SFCG's CVE content or to a control group that viewed non-CVE placebo content that consisted of advertisements from Indonesian entertainment media and retail companies, as well as public service announcement campaigns.

The results indicate that audiences recognized and liked the CVE-themed content at levels comparable to control content, and there were positive effects regarding attitudes toward promoting inclusivity online, although the effect was the result of an unusual, sudden drop in attitudes of control group participants. There also were strong, significant negative treatment effects regarding respondents' attitudes toward living in separated communities.

• Chapter One

  Introduction

• Chapter Two

  Methodology

• Chapter Three

  Findings

• Chapter Four

  Discussion and Recommendations

• Appendix A

  Unnormalized Outcomes

• Appendix B

  Robustness Checks