Adults, Alcohol, & Kids

Mass Media Message

While watching television a few weeks ago I notice that since the new television rating scale has been put into effect advertisers also seem to follow a rating scale that closely resembles that television scale. I have noticed that advertisements tend to be more geared towards the perceived audience during a particular time frame. For example, I have noticed that the majority of television commercials advertising toys and games are generally aired during the early afternoon when children watch cartoons. Commercials for things like cleaning supplies and household goods are aired more during the morning during soap operas and talk shows than at night. Most importantly it seems as if television ads featuring alcohol -- beverages are rarely if ever shown during children and teen programs, and were more likely to be aired during shows that have more adult themes. Children and teen programs seemed to have more nonalcoholic drink commercials than alcoholic drinks commercials. I decided to perform a content analysis of three hours of children and teen television programming and three hours of adult programming to determine if there was a trend.

Hypothesis & Variables

Based on what I thought would occur my research hypothesis was: there would not be an equal number of alcohol and non-alcohol drink commercials during the different types of television programs. My null hypothesis then was: there would be an equal number of alcohol and non alcoholic drink commercials during both children and adult television programming.

I decided to look at two variables. First, I needed to look at drink commercials. This variable was then broken down into two levels. The two levels of my variable were alcoholic and non-alcoholic drink commercials. The second variable that I looked at was type of television programs, in terms of the present rating scale. This variable also had two levels. The two levels were television shows with a rating of TV 14 and lower, and the second level was television shows with a rating of M and above. M standing for mature adult audiences, and TV 14 for audiences over 14.

I chose these variables because they related to what I wanted to analyze. There were no other variables that I could have selected when testing this specific message. However, I could have split the type of television show variable into more than two levels. I could have had a level for each television rating, such as rating G, PG, TV 14, and M. I decided that since all of the ratings of TV 14 and lower were geared towards children and teens they could be grouped together. I was happy with my chosen variables and levels because they were all exhaustive as well as mutually exclusive.

I chose this mass media message because it seemed logical that commercials advertising alcohol would be geared towards adults, and thus advertised more during adult programming than children and teen programs. However, while watching the news a month or two ago it was reported that several congressmen wanted the FCC to make more stricter rules governing when alcohol commercials could be aired. The congressmen felt that some liquor companies had increased advertisements more during children and teen shows to entice underage drinkers than during adult programs. I decided to test and see whether or not alcohol commercials were aired more during children or adult programs.

Coding Scheme & Coding Content

For the message that I wanted to analyze I decided to look at three hours of television programs with a rating of TV 14 and lower and three hour of television programs with a rating of M. When starting my observations I believed that there would be a pattern to the observations. I watched two 1/2 hour shows of Seinfield which was rated PG, 1 hour of 90210 which was rated TV 14, and two 1/2 hour shows of Home Improvement which was rated G. I looked at shows that represented the ratings for children and teen programs.

What I coded was all of the shows' commercials that had an advertisement for either an alcoholic beverage or a non-alcohol beverage. Non-alcoholic beverages were drinks like soda, juice, fruit drinks, kool-aid, punch, milk, sports drinks, and any other kind of drink without any alcohol in it. I also coded all of the drinks that were alcoholic, such as: beer, wine, hard liquor, wine coolers, etc. My coding scheme was basically just keeping count of the total number of alcohol and the total number of non-alcohol drink commercials. I added all of the totals from the three shows together to come up with my totals.

For the shows with a rating of M, I looked at 1 hour of N.Y.P.D. Blue, 1 hour of an episode of Homicide, and 1 hour of Law and Order. Once again I recorded the number of commercials advertising alcohol and the number of commercials advertising non-alcoholic drinks.

Defining each variable as well as each level in the variable was very easy to do. The variable were already predefined, and the levels were also I redefined. The levels in both variables were already defined by society. Either a drink had alcohol in it and was considered alcoholic, or the drink was nonalcoholic. The television variable levels were also predefined. The levels had been predefined earlier last year when the industry came up with the idea of rating television shows.

The variables that I looked at as well as the levels that I coded were easy to define. It was obvious which drinks were alcoholic and which ones were non-alcoholic. For the drinks that were a little uncertain like some wine coolers there were usually words on the bottom of the television screen saying that the drink was alcoholic. The ratings were also very easy because I used the ratings that were already given to each television program, and was shown at the beginning of each television program. The coding scheme was exhaustive because the television variable included all kinds of television programs, and the drink commercial variable also included all kinds of drinks.

The coding scheme of the levels was also exclusive because either the drink was alcoholic or non-alcoholic, and the television shows had a rating of TV 14 and under or a rating of M. None of the observations could fall into more than one category, and there was a category for everything observed. I used a nominal measure for my level of measurement. I placed my observations into different categories.

I used a manifest coding system. The codes that I used to define each variable and each level was very objective, having specific rules and criteria of what counted and what did not count. I used the manifest coding system because it was the most appropriate.

The rules defining which shows had what rating had already been agreed on by the television industry. The shows received the specific rating because of its content. The shows with the M rating was rated for mature audiences because of language, sexual, violent, or inappropriate behavior and other content that was deemed as appropriate only for older audiences such as nudity as was the case in N.Y.P.D. Blue.

The advantages of using a manifest coding scheme was that it was very reliable in terms of what I wanted to code. The manifest coding scheme was very specific and exact. The disadvantage of using this type of coding scheme was that there was little room for flexibility in terms of what I looked at during my observations. For example, I could not count foods like ice cream when during my observation because although it is sort of like a drink it is not exactly one.

I did not use a latent coding scheme because I did not need to look for any underlying meanings. Looking at and coding the surface level features was good enough for what I was observing. The disadvantage of using a latent coding scheme was that it would have not been valid for what I was observing. The advantages of using a latent coding scheme was that I could have assigned my own rating scale to different television show. I would have had a little more flexibility in which category I put the different television shows into. However, I did not need to do this since it had already been done.

My observations were made by making a mark in the correct category for each type of drink commercial and during which program the commercial was aired. While doing my observation and coding the message I did not encounter any problems with my coding scheme or the way that I did my observations.

Chi-Square Analysis

I performed a multiple sample chi-square test. I used two variables each with two levels. The purpose of the chi-square test was to test my null hypothesis, and to see if there was a pattern to my observations or if the observations aligned the way that they did by chance. If the two groups were equal and they aligned a certain way just by chance then the results of my chi-square would prove this.

Variable One: Type of drink commercials
Two levels: Alcoholic/Non-Alcoholic

Variable Two: Type of television show
Two levels: Rating of M/Rating of TV 14 and under

Table of observations

Theoretical Values:

Tl: 85x71/165=36.5
T2: 85 x 94/ 165 = 48.4
T3: 80 x 71/ 165 = 34.4
T4: 80x94/ 165=45.5

Chi-square values:

1. (54-37)2 / 37 = 7.81
2. (31-48)2 /48 = 6.02
3. (17-34)2 / 34 = 289 = 8.5
4. (63-45)2 /45 = 7.2

Total chi-square value = 29.53
DF = 1 Rows (2-1) = 1 Columns (2-1) = 1 1 x 1 = 1
Critical value .05 = 3.84 and .01 = 6.64

The difference between the theoretical and actual observed observations is unusual. Based on the critical values and my chi-square value I can reject the null hypothesis. There is a pattern to the data that I observed. I calculated my chi-square value based on the following formula: chi-square = sum (observed frequencies-theoretical frequencies or value) squared divided by theoretical frequencies.

Interpretation Of Findings

Based on the results of the chi-square analysis there was a pattern to my observations. Television shows with a rating of M have more alcoholic drink commercials than non-alcoholic drink commercials, and television shows with a rating of TV 14 and lower tend to have more non-alcoholic drink commercials than alcoholic drink commercials.

Since the critical value for .05 probability (less than five times out of one hundred) is 3.84 and the critical value for .01 (less than one time out of one hundred) is 6.64, and my chi-square value was 29.53 I rejected the null hypothesis with a 99% confidence level. The observed differences are due to a pattern rather than by chance.

I believe that this pattern was a result of the advertisers knowing which audiences to appeal to in order to sale products. In spite of underage drinkers the majority of the alcohol industry's business would still come from adult drinkers. Also, advertisers know what kinds of drinks appeal to children and this is typically drinks like punch, soda, juice, sport drinks, etc.

Assessment

If I had to conduct this content analysis over, I would break the different television shows up into several different levels like: G, PG, TV 14, and M. I would also break the drink commercials up into more than two levels. The possible levels could be beer, wine, hard liquor, soda, sport drinks, and juice. I could then look at the different kinds of drinks that are advertised during which specific programs. I would then change my research hypothesis to be more specific based on the different kinds of drinks and kinds of television programs. I think that I reamed a lot about advertisements during the course of my content analysis. One thing that I found interesting was that the number of alcoholic beverages advertised in three hours of children and teen programming was much higher than I had expected. I consider 17 commercials advertising alcohol to be a very high number, especially when the target audience is under the legal drinking age.