Out-of-Class Assignment #3
Chapter 9:
1. Distinguish between the following:
a) Internal validity and external validity.
b) Preexperimental design and quasi-experimental design.
c) History and maturation.
d) Random sampling, randomization, and matching.
e) Environmental variables and extraneous variables. a) Validity is defined in experimentation as whether a measure accomplishes what we think previously or not. There are different types of validity, but the two main varieties are internal and external validity. Internal validity answers the question- do the obtained results have a relationship with what we demonstrated in the experiment? Does the experiment show the real cause of what we draw in the conclusion? It is judged by how well it meets the seven internal validity threats. External validity explain if the observed causal relationship can be generalize among persons, settings, and times. It concerns with the relationship between experimental treatment and other factors. It is used when we have larger population.

b) Preexperimental designs are used to control contamination of the relationship that exists between independent and dependent variables. This design is very weak because it fails to control the threats to internal validity or to provide comparison groups that are truly equivalent. Quasi-experimental designs are field experiments that are more advanced than preexperimental. They have control over some of the variables. Using this experiment, we cannot know when or to whom to expose experiment treatment. Also, these designs are not classify as true experiments either.

c) History and maturation are two of the many threats that appear in internal validity. History represents events that occur during the experiment and confuse the relationship studied. Researchers measure those events during the experiment. These events are protest, manipulation, and posttest. In the other side, maturation represents events that occur as a function of the passage of time and are not specific to any particular event.

d) Random sampling is used for researchers that first prepare a sampling frame and then they assign the subjects for the experiment. The sample frame is often small and in this case experimental subjects are recruited. In randomization the experimental group similar to the representatives assigned to the control group. The groups need to be as comparable as possible. Matching is used when randomly assign doesn’t apply. Each experiment and control subject needs to match on every characteristic in order to use this technique.

e) Extraneous variables are variables that influence the relationship between the variables that an experimenter is examining. These variables can be assume or exclude from a research study. Environmental variables are variables that are part of the environment and can affect the way to run an experiment.
3. Why would a noted business researcher say, “It is essential that we always keep in mind the model of the controlled experiment, even if in practice we have to deviate from an ideal model”? Researchers used research methods to determine the factor of causation in a certain event. In order to find the causes of why a specific event under study occurs, researchers keep the model of the controlled experiment. They implement as many variables and techniques as possible in order to reach better results. It is very difficult to execute the ideal model because nobody can develop an experiment and reach ideal results without having any error. I would say that it is almost impossible to do it. For that reason, researchers use the model of the controlled experiment to be as closer as possible from the ideal model.
5. What essential characteristics distinguish a true experiment from other research designs?
One of the main characteristics is that true experiments provide equivalency between experimental and comparison groups. Other research designs fail to provide this. It means that true experiment provides a means by which we can see the equivalence between control groups and experimental using different forms of true experiment. The two designs used in true experiment are pretest-posttest control group and posttest-only control group. Another characteristic is that using these designs, internal validity threats are controlled by random assignment. Also, posttest-only control group design is the only design that can reduced external validity problems. Chapter 10:
1. Distinguish among response error, interviewer error, and nonresponse error. Response error can occur in two separate ways. It can occur when the participant fails to provide the correct answer or it can occur when it fails to provide a complete answer. Basically, this type of error occurs when the record data differs from the true data. These errors occur during coding, editing, or data entry stages. Interviewer error can be caused from a numerous of actions. This occur when the interviewer corrupts the data. Some of the numerous actions that can caused this error are failure to record answers accurately and completely, failure to establish appropriate interview environment, falsification of individual answers, sampling error, physical presence bias, etc. Nonresponse error is one of the factors that can cause bias in interviewing. It occurs when surveys when the responses of participants differ from responses coming from nonparticipants. This can happen when the researcher cannot locate the correct person to be examined or when it is unsuccessful in encouraging the person to participate.

2. How do environmental factors affect response rates in personal interviews? How can we overcome these environmental problems? There are some environmental conditions that affect response rates in personal interviews. They are affected by the location of the interviewer (some neighborhoods are difficult to visit, accessibility of the interviewer (some of them are difficult to reach), time and cost required (these two variables are always together and they are key to develop a good interview), distraction circumstances, and other. Also, questions can be altered or participant coached by interviewers and they can affect the response rates. Some participants do not want to open the door of their houses because they do not feel confidence to answer questions to strangers. In order to overcome these environmental problems, it is very important to have a professionally trained research staff. The manager is in charge to have the adequate people to work in personal interviews. Telephone surveys are useful to have better response rates. Interviewers feel more confident when they are being interviewed by a phone call. Although, the rate of response is not high using this method, you can reduce cost, time and interview bias.
Chapter 11:
1. What can we measure about the four objects listed below? Be as specific as possible.
a) Laundry detergent
b) Employees
c) Factory output
d) Job satisfaction
Research measurement is more controlled and precise than a daily measure of random things. Researches settle for measuring properties rather than the objects themselves. Researchers do not measure properties; they measure indicants of the properties. Below there is a list of indicants that we can measure according with each object. a) Laundry detergent: physical properties that can be measured are volume, weight, color, density, solubility in water, chemical classification according with its properties, etc. Other properties are efficiency of the detergent, efficiency of surfactants, speed of cleaning, and others. b) Employees: we can include properties such as weight, height, age, marital status, sex, nationality, ethnicity, religion. Also, we can measure their skills, job contributions, interest to certain job position, motivation, ambitions, productivity, competitiveness, and others. c) Factory output: we can measure output per week, output per month, output per year, output per resources, volume of output, costs of output, output efficiency, etc. d) Job satisfaction: we can measure employee retention, employee satisfaction survey, applicants per open position, level of job satisfaction survey, healthy work culture, etc.

4. Do you agree or disagree with the following statements? Explain.
a) Validity is more critical to measurement than reliability.
b) Content validity is the most difficult type of validity to determine.
c) A valid measurement is reliable, but a reliable measurement may not be valid
d) Stability and equivalence are essentially the same thing.

a) Validity is more critical to measurement than reliability.
I agree with this statement. Validity gives us the degree to which an instrument measures what it supposed to measure to help the research in solving problems. Reliability has to do with accuracy and precision of certain measurement. A valid measurement is reliable, but a measure is reliable if it provides consistent results. If the results are consistent, it doesn’t mean that they are valid. Reliability contributes to validity, but it is not a sufficient condition for validity. For that reason, validity is more critical to measurement than reliability. b) Content validity is the most difficult type of validity to determine.
I totally disagree with this statement. Content validity is the degree that a measure provides an adequate reflection of the topic under study. If we have a representative sample, then content validity is acceptable. In the case of construct and predictive validity are more complex and abstract because it is relate with our ability to predict and estimate outcomes. c) A valid measurement is reliable, but a reliable measurement may not be valid.
This statement is correct. By definition a valid measurement is reliable, but a reliable measurement not always is valid. Reliability is a contributor to validity, but a measurement can be reliable without being even valid. We need to have consistent results in order to consider a reliable measurement as valid. d) Stability and equivalence are essentially the same thing.
I disagree with this statement. Stability is when a measure secure consistent results after repeated measurements are applied. All the measurements are applied to the same person or object. It concerns with personal and situational fluctuations from one time to another. In the other side, equivalence concerns with variations in results after measure different persons or objects. It compares results among objects or person and see if the results are the same. In conclusion, stability is same results from the same object or person and equivalence is same results from different objects or persons.