Chapter 2: Elements of Research PDF

CHAPTER 2 Elements of research CONCEPTS AND CONSTRUCTS  A concept is a term that expresses an abstract idea formed by generalizing from particulars and summarizing related observations.  For example, a researcher might observe that a public speaker becomes restless, starts to perspire, a...

CHAPTER 2 Elements of research CONCEPTS AND CONSTRUCTS  A concept is a term that expresses an abstract idea formed by generalizing from particulars and summarizing related observations.  For example, a researcher might observe that a public speaker becomes restless, starts to perspire, and fidgets with a pencil just before giving a speech. The researcher might summarize these observed patterns of behavior and label them “speech anxiety”.  A conceptual definition defines a word or a concept by substituting other words or concepts for it. CONSTRUCTS  A construct is a concept that has three distinct characteristics.  First, it is an abstract idea that can be broken down into dimensions represented by lower level concepts.  Second, because of its abstraction a construct can not be observed directly.  Third, a construct is usually designed for some particular research purpose so its exact meaning relates only to the context in which it is found.  For example, the construct involvement has been used in many advertisement studies. it is a construct that cannot be observed directly, and it includes concepts of attention, interest and arousal. In some context involvement means a subject involvement with the product, in others it refers to involvement with the message or the medium. Its precise meaning depends on the research context. VARIABLES – The empirical counterpart of a concept or a construct is called a variable. Variables are important because they link the empirical world with the theoretical. – Variables refer to the different phenomena, concepts and events that are measured, manipulated and tested in research. – Examples are: Facebook usage, TV viewership, trust in media, credibility of sources, attention to certain media messages, etc. – Variables can have more than one value along a continuum. Each variable can take different values or degrees. (a lot, little, or not at all – high, moderate or rare, etc.)  For example, the variable “satisfaction with pay per view TV programs” can take on different values- a person can be satisfied a lot- a little- or not at all – reflecting in the empirical world what the concept “satisfaction with pay per view TV programs” represents in the theoretical world.  Researchers try to test a number of related variables to develop an underlying relationship among them TYPES OF VARIABLES: 1: INDEPENDENT AND DEPENDENT VARIABLES:  Variables are classified in terms of their relationship with one another – into independent and dependent variables.  Independent variables are varied by the researcher; dependent variables are observed and their values are presumed to depend on the effect of the independent variables. In other words, the dependent variable is what the researcher wishes to explain.  That’s to say, when a change happens in the value of the independent variable the values of the dependent variables will change accordingly. - Some examples are: Viewing television violence  Aggressive behavior Using online news  political participation Exposure to TV advertising  body image  The distinction between types of variables depends on the purpose of the research. An independent variable in one study may be a dependent variable in another.  A research task may involve examining the relationship of more than one independent variable to a single dependent variable – a researcher could investigate the effect of camera angles and newscaster style on his credibility.  Moreover, multiple dependent variables are measured in their relationship to one independent variable in a single study which is called multivaiate analysis. 2: DISCRETE AND CONTINUOUS VARIABLES  Two forms of variables are used in mass media investigation. A discrete variable includes only a finite set of values; it cannot be divided into subparts.  For example, the number of children in a family is a discrete variable because the unit is a person. Other examples include political affiliation, gender. A continuous variable can take on any value including fractions and can be broken into subsections.  Time spent watching television is an example of continuous variable. It is perfectly meaningful to say that Person A spent 3.12115 hours while Person B watched 3.12114.  When dealing with continuous variables, researchers should keep in mind the distinction between the variable and the measure of the variable.  For example, if a child’s attitude toward violence is measured by counting his positive responses to6questions, then there are only seven possible scores: 0,1,2,3,4,,5,6. the underlying variable is continuous even though the measure is discrete.  As a generalization, most of the measures in mass media research tend to be discrete measures of continuous variables. OTHER TYPES OF VARIABLES:  In non experimental research where there is no active manipulation of variables different terms are sometimes substituted for independent and dependent variables.  The variable that is used for predictions or is assumed to be causal is called the predictor or antecedent variable.  The variable that is predicted to be affected is sometimes called criterion variable  Researchers often wish to control certain variables to eliminate unwanted influences. These control variables are used to ensure that the results of the study are due to the independent variables, not to another source.  On occasion, researchers use a control variable such as age, gender or SES to divide subjects into specific relevant categories.  For example, in studying the relationship between newspaper readership and reading ability, researcher know that IQ will affect the relationship and must be controlled; thus subjects may be selected based on IQ scores or placed in groups with similar scores. DEFINING VARIABLES OPERATIONALLY:  An operational definition specifies the procedures to be followed to measure a concept/ variable.  It is a clear statement of what is to be observed/investigated to identify the values of a variable. - Operational definitions are indispensible in scientific research because they enable researchers to measure relevant variables. - In any study, it’s necessary to provide operational definitions for both independent and dependent variables.  Kerlinger identifies two types of operational definitions, measured and experimental.  A measured operational definition specifies how to measure the variable  For example, in examining the relationship between facebook use and envy and depression among college students, envy was operationally defined by the subject's score on an 8 item envy scale.  An experimental operational definition explains how an investigator has manipulated a variable. This type of definition is used when the independent variable is defined in a laboratory setting.  For example, in a study on the impact of television violence, the researcher might manipulate media violence by constructing two 8 minute films. The first labeled “the violent condition” could contain scenes from a boxing match. The second labeled “the nonviolent condition” could depict a swimming race.  Operationally defining a variable forces a researcher to express abstract concepts into concrete terms; communicating exactly what the term represents.  Some variables are vague or ambiguous and need to be redefined in a measurable way.  Because operational definitions are expressed so concretely, they can communicate exactly what the variables represent. -For example, a researcher might conceptually define political knowledge as factual information about politics and government that individuals retain in their memory; and operationally define it as the number of correct answers on a 20 item true/false test. Consequently, there is no confusion as to what the statement “women posses more political knowledge than men” actually means.  Finally, there’s no single method for operationally defining a variable, each researcher must decide which method is best suited for the research problem at hand. LEVELS OF MEASUREMENT  Scientists have distinguished four different ways to measure things or four different levels of measurement, depending upon the rules that are used to assign numbers to variables. Based on these levels, statistical procedures that can be conducted within the variables are determined. The four levels are nominal, ordinal, interval and ratio. THE NOMINAL LEVEL  The nominal level is the weakest form of measurement.  In this level, numerals or other symbols are used to classify people, objects or characteristics.  Numerals are simply labels that stand for respective categories; they have no mathematical significance  An example of nominal level in mass media is classifying respondents according to the medium they depend on most for news. Those depending on TV may be category 1, those depending most on newspapers may be category 2, those depending on magazines in category 3.  Another example may include gender; males may take number 1, females the number 2. the number 1 did not signify a lower quantity; instead, the numbers 1 and 2 were used to classify respondents into categories of males and females. CHARACTERISTICS OF THE NOMINAL LEVEL: The nominal level is characterized by equivalence. If an object is placed in category 1, it is considered equal to all other objects in that category.  all categories are exhaustive and mutually exclusive (‫)جامعة مانعة‬. This means that each measure accounts for every possible option and that each measurement is only possible for only one category. Example: The variable of gender consists of 2 categories males and females. These are exhaustive categories as there is no third option. Additionally, it is mutually exclusive since each subject falls in one category only and can not be placed in more than one category at the same time. Other examples include race, eye color, political party. ORDINAL LEVEL:  Objects measured at the ordinal level are usually ranked along some dimension, such as from smallest to largest.  For example, one might measure the variable “socioeconomic status” by categorizing families according to class: lower, lower middle, middle, upper middle or upper. A rank of 1 is assigned to lower, 2 to lower middle, 3 to middle and so forth. In this situation the numbers have some mathematical meaning: families in category 3 have higher SES than category 2.  Note that nothing is specified with regard to the distance between any two rankings.  Examples of ordinal measure in media research include classifying subjects into categories based on the hours spent watching TV or surfing the internet. Each group will be given a particular rank according to the hours spent using the medium. Subjects will be classified from smallest to largest (or vise versa) regardless of the difference in the amount of time spent between groups. In other words the difference between group 1 and 2 may be one hour, whereas the difference between group 2 and 3 may be 45 minutes and so forth.  Other example includes income level (“less than 50K”, “50K-100K”, “over 100K”). CHARACTERISTICS OF ORDINAL LEVEL:  An ordinal level posses the property of equivalence. Thus in the previous example, thus all families placed in a category are treated equally even though some might have greater income than others.  It also posses the property of order among categories. Any given category can be defined as being higher or lower than any other category. INTERVAL LEVEL  When a scale has all the properties of an ordinal scale and the intervals between adjacent point on the scale are of equal value, the scale is at the interval level.  The most obvious example of an interval scale is the temperature. The same amount of heat is required to heat the object from30 to 40 degrees as to warm it from 50 to 60 degrees. Interval scales incorporate the formal property of equal differences.  One disadvantage of the interval scale is that it lacks a true zero point, or a condition of nothingness. For example, it is difficult to conceive of a person of having zero intelligence or zero personality.  The absence of a true zero means that a researcher can not make statements of a proportional or comparative nature. For example, someone with an IQ of 100 is not twice as smart as someone with an IQ of 50, and a person who scores 30 on a test of aggression is not 3 times as aggressive as a person who scores 10.  Examples of interval level in media research include the amount of time spent using a particular medium, respondents’ ages, income RATIO LEVEL  Scales at the ratio level have all the properties of interval scales plus one more: the existence of a true zero point.  With the introduction of this fixed zero point, ratio judgments can be made.  For example, since time and distance are ratio measures, one can say that a car travelling at kilometer per hour is going twice as fast as a car travelling at 25 km/h.  Ratio scales are relatively rare in mass media research. Zero means the absolute absence or non existence of the phenomenon which is not the case in humanities  Variables measured at the nominal and ordinal levels are always discrete variables.  Variables measured at the interval or ratio levels can be either discrete (as number of TV in a household) or continuous (as number of minutes spent per day reading magazines) MEASUREMENT SCALE  A scale represents a composite measure of variable; it is based on more than one item.  Scales are generally used with complex variables that do not easily lend themselves to single indicator measurement  Some items such as age, number of radios in the house can be adequately measured without scaling techniques.  Measurement of other variables such as attitude toward TV news or gratification from going to a movie require the use of scales. LIKERT SCALES  The most commonly used scale in mass media research is the Likert scale.  A number of statements are developed with respect to a topic and respondents can strongly agree, agree, be neutral, disagree or strongly disagree with the statements.  Each response option is weighted and each subject’s responses are added to produce a single score on the topic. THIS IS THE BASIC PROCEDURE FOR DEVELOPING A LIKERT SCALE 1- compile a large number of statements that relate to a specific dimension. Some statements are positively worded and some are negatively worded. 2- administer the scale to a randomly selected sample of respondents 3- code the responses so that the high scores indicate stronger agreement with the attitude in question. 4- analyze the responses and select for the final scale the statements that most clearly differentiate the highest from the lowest scores SEMANTIC DIFFERENTIAL SCALES  Another commonly used procedure is the semantic differential technique. It is commonly used by communication researchers as a measure of attitude.  To use this technique, a name or concept is placed on top of a series of seven point scales anchored by bipolar attitudes SIMPLE RATING SCALES  Rating scales are commonly used in mass media research. Researchers commonly ask respondents to rate a list of items such as to rate how much respondents like radio or TV on air personalities.  The researcher’s decision is to decide on which type of scale to use: 1to3, 1to5, 1to7…. Generally, there are few things to consider: 1- a scale with more points rather than fewer points allows for greater differentiation on the item being rated. 2- experience show that males and females, all age groups like to use a 1to10 scale. Everyone understands a 1to 10 scale where a 10 is the best or perfect and 1 is worst or incorrect 3- when using simple rating scales, it is best to tell respondents that the higher the number the more you agree or the higher the number the more you like.