COMM 3100 Exam 2

What is a difference between a variable and a concept? What makes a variable a variable?

a. A concept is simply an idea. A concept can be (and usually is) made up of many different variable. A variable is one thing - a noun—that varies, that can have different levels or values.

What is the difference between moderating and mediating variables?

a. A moderating variable is a third variable that causes both the independent and dependent variables to occur (or, at least, as has effect on the iv and dv). A mediating variable is a variable that, when caused by the independent variable, in turn causes the dependent variable to occur (in other words, iv causes med, which causes dv).

Explain the two types of relationships specified in a hypothesis.

a. A relationship of association, where one variable is found, the other will be found also. If 'a' then 'b'. b. A relationship of cause ("causal relationship") implies that one variable causes change in the direction of the other variable, if 'a' does this, then 'b' does that. ('a' leads to 'b,' 'a' causes 'b,' 'a affects 'b,' 'a results in 'b,' etc.)

What are reversible and irreversible relationships between variables?

a. A reversible relationship can go either way when asking the question 'which came first'. b. An irreversible relationship can only go only one way ('a' must always go before 'b.')

What is a representative sample?

a. A sample is a group of people chosen to be included in the study as participants. The sample is expected to be representative of the population under study. A representative sample thus "stands in for" the population without actually having to interview the entire population. A sample may be representative in measurable terms, such as in quantitative research, or in descriptive terms, such as in qualitative research.

What is a hypothesis?

a. A statement (prediction) the researcher makes about the relationship between two variables, specifying how two concepts are believed to be related. It is an educated guess about what will happen in a relationship between variables.

Is a directional hypothesis preferred over a non-directional hypothesis? Is one easier to study than the other?

a. Both are acceptable. Whether a hypothesis is directional or non-directional depends on previous research that is available. It may be possible that enough literature can support a directional hypothesis.

What limitations are associated with using a non-random sample in your study? For example, if you only asked survey questions to a communication class instead of students across the entire university.

a. Cannot generalize data to a larger population. Sample may be too small to determine a statistically significant relationship if one occurs.

What would you label class standing as? Would it be nominal or ordinal?

a. Class standing (freshmen, sophomore, junior, senior) are variables that can be ranked, and would therefore be categorized as ordinal.

Why is conceptual fit so important?

a. Conceptual fit refers to how closely the way we define our variable and the way we measure that variable, are. It's important to preserve the conceptual fit in a study, because it's important to measure what we say we are measuring.

How do history, maturation, testing and instrumentation affect internal validity?

a. History - things that occur outside the study that affect the results within the study. These external factors may influence how participants behave or respond in a study. b. Maturation - this deals with changes within the participants themselves over time. These changes can be psychological or physical. c. Testing - if a participant is more aware of or more comfortable with the testing procedure (like questions on a survey), they may respond differently than anticipated. d. Instrumentation - giving an alternate second test may produce different results simply because of the nature of the second test. It may not establish a true comparison between the first test and the second test.

What are the four levels of measurement?

a. Nominal- makes use of unordered categories, classifying the variable into qualitatively different and unique categories; for example, political party affiliation and religious affiliation. b. Ordinal- ordered category or rank, we can determine whether an observation is greater than, less than, or equal to other observations. But it does not indicate the how much the difference is; examples are top 5 college basketball teams or tv programs classified as high, medium, or low in violence. c. Interval-specifies relative positions and also establishes standard equal distances between points on a scale but has no meaningful zero point; for example, a 5-point likert scale. d. Ratio-has a true meaningful zero point, measured in equal intervals, and can be mathematically measured; for example, a heart rate, years at a current address, income, or driving speed.

Is a directional hypothesis one tailed or two tailed? Explain why.

a. One tailed, the hypothesis is going in one specific direction.

List three types of sampling used in qualitative research.

a. Purposive sampling (for example, seeking out pregnant students on campus) b. Quota sampling (for example, non-randomly asking a certain number of freshmen with declared majors and undeclared majors as well as a certain number of females and males within each group) c. Maximum variation sampling (for example, if studying student organization communication, you would pick different-sized organizations with varying distributions of class ranks, gender, and ethnicities).

What is the difference between reliability and validity?

a. Reliable measures are able to produce the same results when replicated. A reliable measure is considered consistent and stable. Valid measures are accurate in terms of measuring the constructs under study. A study is considered valid if it is proven to study what it is intended to study.

What are some examples of the different types of relationships between variables?

a. Reversible - if we hypothesize that the purchasing of violent video games by children co-occurs with violent behavior (without predicting which causes which), we are predicting a reversible relationship. Irreversible - if we hypothesize that increased fat intake results in obesity, but not vice versa, we are predicting an irreversible relationship. b. Deterministic - rare in communication since conclusions in our field often cannot be made with such certainty. But this relationship may be seen in scientific hypotheses, such as the prediction that a certain level of body fat always results in higher likelihood of heart disease. c. Stochastic - if we predict that increased fat intake may not be directly related to obesity, but does some of the time, we are predicting a stochastic relationship. There is a sense of probability since one could take in one high-fat meal a day but not gain weight, while another person may gain weight eating several large, low-fat meals in one day. Stochastic is more common in communication since the relationship may not work every single time in every single situation. d. Sequential - if we predict that increased exposure to positive comments about self, followed by a good grade on a test, leads to increased positive self-image, we are predicting a sequential relationship. Co-extensive - if we predict that determining a speaker's authority and intent can occur at the same time when establishing credibility, we are predicting a co-extensive relationship e. Sufficient - if we predict that increased studying -by itself—will result in getting a good grade on a test, we're predicting a sufficient relationship. Contingent - if we predict that increased studying has to be paired with increased class attendance in order to get a good grade on a test, we are predicting a contingent relationship. f. Necessary - if we say that communication cannot occur without a decoded message, we are predicting a necessary relationship. Substitutable - if we say that communication can occur with either verbal or nonverbal messages, we're predicting a substitutable relationship.

What is the difference between your sample and your sample frame?

a. Sample frame is the specific part of a population that you have access to. The sample are those you chose to participate in the study within that sample frame.

What are four types of sampling used in quantitative research?

a. Simple random sampling b. Stratified sampling c. Proportional stratified sampling d. Cluster sampling

What does a confidence level of 95% mean?

a. That there is a 95% probability that your answer/result is pretty close to correct.

Explain the independent and dependent variables.

a. The IV is the variable that is being applied to the dependent variable. Changes in the IV causes changes in the dv. The dv is the variable that is being observed for any changes if the IV is being changed. b. Example: gender and use of humor in communication. In this case, gender is the iv and use of humor is the dv

What is usually the minimum acceptance level of reliability?

a. The conventional criterion is that a coefficient of about .70 is the minimal acceptance for level of reliability.

Why is sample size important to consider when you want to have generalizability?

a. The larger the sample, and the more random it is (meaning, that each member of a population has equal chance of being selected).

Explain a null hypothesis.

a. The null hypothesis simply says the research hypothesis is wrong, that there is no (null) relationship between the variables that the research predicated. The null hypothesis is implied after stating the proposed hypothesis and does not need to be written. If the original hypothesis is not determined to have a certain probability of being correct, then the null hypothesis is usually accepted. A research hypothesis is always contrasted with a null hypothesis. The null hypothesis is symbolized by a capital h and followed by the subscript zero (h0).

What are the different types of operational definitions?

a. There are two types of operational definitions: measured operational definitions (how a researcher can measure the existence or quantity of a variable) and experimental (or manipulated) operational definitions (how a researcher can manipulate a variable in an experiment).

What is the difference between a uni-dimensional and multi-dimensional variable?

a. Uni- dimensional variables have only one component or dimension, or only one way of looking at them. Multi-dimensional variables embody more than one component or dimension, or, in other words, have multiple ways of looking at or measuring them.

What is a random sample?

a. When each person in the population has an equal chance of being chosen for the sample. A random sample can be generalized to the population.

How do I know when I have achieved a great sample for research?

a. When the sample is representative of the overall population you are studying.

What is an example of a study that uses (random) cluster sampling?

a. When wanting to look at communication between employees and their superiors in a large company, you may not have access to all departments. So, to overcome this, you collect data from selected departments you do have access to. The data collected may not be generalizable to the whole company.

How do I know if my measures are valid?

a. You can start by checking its face validity by asking if your procedures make sense at face value. Then you could look at other used measures and see how yours holds up when compared to other existing measures. You should also consider construct validity and decide whether there are logical relationships between your variables.

How can I ensure that my measurement tool is reliable?

a. You can test-retest by giving the same measure to the same people at two different times. You can implement an alternate form of the measure to find out if the order of the items in your measure are affecting the ways people respond. You can split-half the items with a randomizing tool. To ensure that you and others participating in your research procedures are recording data in the same way you can test for inter-coder reliability.

What's the difference between conceptual and operational definitions?

a. Your conceptual definition will define the concept or variable that you are studying. Often considered a written concrete concept, a conceptual definition is a scholar's definition of the variable being studied (usually based on previous research). Your operational definition—which describes the observable characteristics of a concept that can be measured, identified, or observed-- will describe how you will plan to measure and/or observe the concept or variable of interest.