factory to your food store, it naturally settles down; and when opened, there may be some air space at the top.

Many of the products in bottles or cans are processed at extremely high temperatures. During the cooling period, the contents contract, and this naturally leaves "head space" at the top of the container. There are other practical reasons why products may have air space at the top when the container is opened. For example, some foods are usually spooned from wide mouth jars, and if filled to the top would cause spillage when used.

However, the package must always contain the full amount of the net contents specified on the label.

YOU CAN HELP IMPROVE LABELING

Now you know some of the practical reasons for certain aspects of packaging and the grocery manufacturers' interest in providing good descriptive labels.

With the help of thoughtful consumers, the pattern for good descriptive labeling has been established. Grocery manufacturers want to know what you like and what you don't like about the labels on their packages because such knowledge helps them to provide the kind of information on labels that will be most helpful to you.

This is what grocery manufacturers would like you to do:

Read the labels on grocery packages and learn all about the products.

Talk about labels with your friends and encourage them to read them.

If you like something about the label of a particular product, or have suggestions for changes or additions, write the manufacturer. If you want additional information about a product which is not shown on the label, write the manufacturer.

The Grocery Manufacturers of America, Inc. will be pleased to have your comments on labels generally and any suggestions for improvement.

Grocery Manufacturers of America, Inc.

205 East 42nd Street New York, New York 10017

Paul S. Willis, President

YOUR RIGHT TO CHOOSE AMONG COMPETING PRODUCTS IS YOUR DEPENDABLE GUARANTEE OF FREEDOM.

Senator NEUBERGER. Do you want Mr. Barlow next?

Mr. WILLIS. Yes.

STATEMENT OF WALTER BARLOW, PRESIDENT, OPINION

RESEARCH CORP., RESEARCH PARK, PRINCETON, N.J.

Mr. BARLOW. My name is Walter Barlow and I have been with Opinion Research Corp., Research Park, Princeton, N.J., since 1946 and since 1960 have been its president.

My first position with ORC was as a researcher on the Public Opinion Index for industry; and in 1948, was named managing director of the Index and vice president of ORC. Later I became executive vice president.

Opinion Research Corp. was founded in 1938. Our fields of research activities include product, consumer, and marketing research; public relations, corporate communications programs, and investor relations; personnel and industrial relations and management development; and the mass media.

ORC clients include companies, associations, governmental agencies, and nonprofit groups. Over the years we have conducted research for more than 350 organizations, about 100 of which maintain a continuing relationship with us.

Our Princeton headquarters staff has more than 100 members including psychologists, sociologists, statisticians, marketers, economists, and specialists in business administration. In the field, and by that I mean throughout the country, we maintain a permanent interviewing staff of 1,200 men and women.

Through the ORC national probability sample, we can obtain an accurate cross section of attitude and behavior relating to virtually any subject.

PURPOSE

The research assignment given to Opinion Research Corp. by the Grocery Manufacturers Association, is delineated in the statement of purposes:

1. To determine the attitudes of American consumers toward food and grocery manufacturers and their products.

2. To determine consumer attitudes about grocery packaging, with special emphasis on, as Mr. Willis mentioned:

Package shapes and designs.

Range and variety of package sizes.

Degree of package fill.

Information on packages and labels.

Statements of number of servings.

Food product illustrations on packages and labels.

SURVEY DESCRIPTION

Findings reported herein are from a coast-to-coast survey employing national probability samples of female and male heads of household age 18 and over. All interviews were personal interviews, made by interviewers trained in probability sampling technique.

Advanced sampling methods were used to insure the greatest possible validity and reliability of the findings, and to insure their projectability to the total American population of household heads, female and male.

Instructions to the interviewers predesignated mechanically both the households to be included in the study and the specific individuals to be interviewed, precluding interviewers from making such choices. If the designated respondent was not available at the first call, interviewers made at least three additional calls attempting to interview him or her.

The sample of 1,420 women provides a scientific cross section of 51 million female household heads, almost all of whom are responsible for buying and preparing food for home consumption.

For purposes of interviewing, the "female household head" is defined as "the lady of the house"-whether or not a male head of the household was also in residence.

The sample of 505 men provides a cross section of the 45 million heads of households, defined as "the man of the house." Though not as active shoppers as the women, the men interviewed evidenced lively interest in the subjects covered in this survey.

PRIMARY SAMPLING UNIT

The entire area of the United States was divided into about 1,700 primary sampling units. With a few minor exceptions, a primary sampling unit consists of a county or a group of contiguous counties. All primary sampling units were allocated to 86 what we call strata. Each stratum consisted of a set of primary sampling units as much alike as possible with respect to such characteristics as geographical location, size of central city, rate of population growth, and economical characteristics.

Because of their size, 22 large metropolitan areas were in strata by themselves and were automatically included in the sample as "selfrepresenting" areas. One sample area was selected in a random manner from each of the remaining 64 strata. Within a stratum the probability of selection of any one primary sampling unit was proportionate to its estimated population.

INTERVIEWING LOCATIONS

A total of 240 sample segments, or interviewing locations, was allocated to the 86 sample areas. Sample segments were small land areas which included an estimated minimum of 15 housing units.

In urban areas sample segments were defined as blocks or groups of blocks; in rural areas sample segments were defined by recognizable boundaries such as roads, streams, and other distinct landmarks.

Within a sample area the probability of selection of a sample segment was proportionate to its estimated population. Several sources were used in estimating the size of each segment. These included Bureau of the Census block statistics and mapping materials as well as special field visits-on the part of our staff-to subdivide census enumeration into segments.

HOUSEHOLDS

For each sample segment the interviewer was provided with a detailed map and instructions for listing a specific group of households in which interviews were to be conducted. No deviation from the specified procedures was permitted.

SPECIFIC INDIVIDUALS

Within each sample household, the specific person to be interviewed either head of household or wife of head was predesignated. To attain maximum control of nonresponse, interviewers made up to four calls at households where persons were not at home on the interviewer's first or subsequent calls.

SAMPLING RELIABILITY

The methods employed in conducting this study assure close control over each stage of sample selection, as we have been describing. Even with a refined sampling method, however, research results are subject to some variation arising from the fact that the findings are based on a sample rather than upon interviews with all members of the survey population.

Approximate magnitudes of sampling tolerances for reported survey percentages are shown in the table below. These sampling tolerances estimate the limits, in percentage points, of possible sample variations at the 95 percent confidence level.

Approximate sampling tolerances for percentages

For instance, of the 1,420 women interviewed 31 percent rated their impression of food manufacturers "very favorable." Referring to the top line of the table, third column, chances are 95 in 100 that the "true" percent for all 51 million female household heads is between 28 percent and 34 percent.

The results of the study are reported in the booklet "What Consumers Think," which has been submitted with Paul S. Willis' testimony. With your permission I shall now briefly review the results as reported in this booklet.

If I might refer to pages 2 and 3. It has been covered in my previous statement, so we might turn to page 4 and continue through the summary pages that are the next four.

Here is a summary. First of all regarding food manufacturers, American consumers, both men and women, have a favorable view of food manufacturers and what the manufacturers do for them. Food manufacturers are looked upon more favorably, both in general and on most specifics, than the other consumer goods industries studied for comparison purposes.

Strong points in consumers' opinions of food manufacturers center around what they do for the consumer-such as providing a wide variety of good, safe, dependable products, well packaged, and easy and convenient to buy and to prepare.

Quality is of high importance to the industry. There is a pattern of satisfaction and enthusiasm for the quality of food and other grocery products offered to consumers today.

Most people are highly satisfied with the quality of products and their nutritional value. They are convinced that they are clean and safe to eat and they praise their ease of preparation and use.

An exception to the very favorable attitude pattern relates to "dangerous substances" in food.

While most consumers state that there are no substances in today's packaged food that are dangerous to people's health, there are some who think that such substances are present.

As relating to packaging: In general, consumers are well pleased with the packaging of food and other groceries.

Both men and women appreciate the progress that has been made in packaging, particularly in the new and improved packaging techniques and materials and the greater convenience of handling, storage and use of products.

Most express satisfaction with the specific aspects of grocery packaging studied-convenience, range of sizes, information printed on labels and packages, food illustrations, statements of number of food servings, and degree of fill.

There are, however, some minority dissatisfactions relating to grocery packaging.

As to reactions to proposed changes in packaging practices: One of the purposes of the research was to investigate reactions to proposals that changes be made in packaging practices. Each proposal was presented to the respondent together with a paragraph summarizing the industry point of view on the issues.

The respondent was then asked to give his opinion as to whether changes are needed, what changes, and who should determine what changes are made.

After hearing the proposals for changes in packaging and the manufacturers' viewpoints on these changes, most consumers express satisfaction with present industry practices.

While a few feel a need for change in some practices, fewer than 1 in 19 are in favor of government agencies determining what changes should be made.

You indicated a shortage of time, Madam Senator. Perhaps I should pause at this point and let the questions come. To go through the balance will probably take more time than you would care to allow. Senator NEUBERGER. Senator Dominick?