Violence, sex and profanity in films: correlation of movie ratings with content.

Abstract

CONTEXT: Children's exposure to violence, sexual themes, profanity, and the depiction of substances in movies remains a source of parental and public health concern. However, limited research quantifies the correlations between movie content, ratings, and economics or addresses the issue of ratings "creep." OBJECTIVES: To characterize available information about violence, sex, and profanity content of movies as a function of rating; quantitatively explore the relationships between content, ratings, and economic information; compare the amount of violence in animated and non-animated G-rated films; and test for a trend of decreased stringency of rating criteria (ie, "ratings creep") as a function of time. DESIGN: We developed a complete database of movie ratings available from the Motion Picture Association of America (MPAA) to characterize the content information (including any indicated reasons noted for ratings) for all movies released between January 1, 1992 and December 31, 2003. We then added to the database the three Kids-in-Mind content-based scores for: (1) violence and gore, (2) sex and nudity, and (3) profanity; and the 15 categories of information from Screen It!, which began providing information in mid-1996. Finally, we obtained information on gross revenues and movie budgets from the IMDbPro. We performed statistical analyses to correlate the content-based scores with the overall rating and rating reasons assigned by the MPAA; to test the hypothesis that age-based ratings became less stringent over time; to explore correlations between film content, ratings, and available economic information; to compare the amount of violence in animated and non-animated G-rated films; and to characterize the available information about the depiction of substances in films. MAIN OUTCOME OF MEASURES: Description of movie ratings, correlation of content with rating, and statistical results. RESULTS: Comparing the content-based scores for different movie ratings, we find large variability exists in the types of content that receive different MPAA ratings, and good correlation between the content-based scores assigned by Kids-in-Mind and Screen It! The MPAA rating reasons correlate with higher scores assigned to content-based ratings, and the number of reasons indicated increases with the age-based rating category. We found significantly higher rated content in movies as a function of time, suggesting that the MPAA applied less stringency in its age-based ratings over time for the period of 1992-2003. Animated films rated G by the MPAA received a significantly higher content-based score for violence on average than non-animated films rated G (P < .05). With respect to information about the depiction of substances, the MPAA mentioned alcohol or drugs in its rating reason for 226 films (18%), while Screen It! identified depiction or use of tobacco, alcohol, and/or drugs in 1211 films (95%), including 26 of the 51 G-rated films (51%). We found significantly higher gross revenues for PG-13- and R-rated films when comparing films that received an MPAA rating reason for violence compared with those films that did not (P < .001 based on 2-sided t-tests with unequal variances for both of the separate tests of PG-13- and R-rated films). CONCLUSIONS: Parents and physicians should be aware that movies with the same rating can differ significantly in the amount and types of potentially objectionable content. Age-based ratings alone do not provide good information about the depiction of violence, sex, profanity, and other content, and the criteria for rating movies became less stringent over the last decade. The MPAA rating reasons provide important information about content, but they do not identify all types of content found in films and they may particularly miss the depiction of substances.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 60 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.