Measuring the Construct of Passivity, Research Paper Example

Developing Norms and Standardizing Assessments       

In order to develop a self-report test for the psychological construct of passivity, it is first necessary to determine the norms for the test. Assuming that the test for passivity is going to be used to test adults, the norms for the test can be measured by available criteria. There are already established intelligence and cognitive norms for assessing passivity, so the test can measure the responses of the test subjects against these intelligence and cognitive norms above 30% of normal distribution.  A z-score can be used to determine how far the individual respondent diverges from the mean and the standard error of measurement. A z- score of 1.0 would represent 84.13% percentile whereas a z-score of -1.0 would represent 13.57% percentile. In that range, it is good way to find out if the hypothesis in question is valid or null.

It is also necessary to ensure standardization for the test, which means that the test must be administered in the same basic environment and under the same conditions. In practical terms this means that the questions in the self-report assessment must be the same for all subjects; that the subjects must be from the same basic demographic groups; that the test must be administered in the same manner for all subjects; and other psychological conditions and physiological factors must be standardized according to the approved psychological testing manuals. By ensuring standardization in the development and administration of the test, it is thereby ensured that the source of the test can be as accurate as possible.

Multidimensional and Unidimensional Assessments

Because passivity is largely focused on how people relate to others and whether they take an active or passive role in dealing with conflict. Multidimensional assessments contain180 factors of intelligence and cognitive processes or three-dimensional model that are supposedly to evaluate three kinds of cognitive ability: product, operations, and content. The main focus is to evaluate the items that are commonly marked by the groups with like-mindedness of passivity.  A unidimensional assessment examines a single factor, even though multiple questions may be asked in order to triangulate the responses. A single question, such as “how do you respond when someone is angry with you?’ might yield a response that indicates passivity, but it is too general to offer much insight. For example, a person who responds passively to his or her confrontational boss at work might be more assertive in responding to an angry spouse or romantic partner at home. This does not mean that multiple questions are always a means of constructing a multidimensional assessment, as multiple questions can still be used to assess a specific, single factor. If a series of multiple questions is intended to determine whether a person is passive in the workplace, then that could be seen as a unidimensional assessment even though there are a variety of questions.

With this in mind, a multidimensional assessment of passivity could include questions that focus on a variety of scenarios. Asking questions about how an individual responds to various people in various scenarios would be multidimensional. Such questions could include asking about responses to specific scenarios involving friends, family members, coworkers, romantic partners, authority figures (such as police) and other people and scenarios where potential conflict is present.

Test-Retest Reliability

Test-retest reliability is related to the correlation between overall responses when a test is administrated more than once. Determinations made about personality traits or psychological constructs such as passivity are presumably factors that would not differ greatly from one test to another. In order for test-retest reliability to offer insight, the conditions in which the assessments are given must be duplicated as closely as possible during each assessment. Along with the duplicated conditions, the questions must also be the same, otherwise the differences in questions would skew the results. Passivity is a psychological construct that is not likely to change significantly from one test to another, so it would be expected that a subject who demonstrates the attributes of passivity in one test would demonstrate similar responses or better in a second test. If the responses in both tests show closely related to the scores or correlation shows closer to plus or minus 1.0, that can be seen as evidence that the scatterplot chart and correlational estimates are reliable to score the validity of the test.

Where test-retest reliability measures the correlations in responses between two or more administrations of the same assessment, internal consistency reliability relates to the questions within the assessment. If there are several questions that are designed to focus on the same issue, but are worded differently, it would be expected that the subject would provide similar responses. For example, a may indicate that he or she disagrees with the statement “when confronted by my boss, I typically choose to speak my mind.” The same subject could then be asked to agree or disagree with the statement that “I prefer to keep quiet when my boss tells me I did poorly on a project.”  If the subject agrees with this statement, this shows that there is good internal reliability between the questions in the assessment, as both responses indicate that the subject responds passively in confrontations with his or her boss.

Internal Consistency Reliability

With this in mind, it is important to develop questions and statements for the assessment on passivity that allow for measuring internal consistency reliability. A single question about a particular scenario or issue may seem to indicate that the subject is passive from a psychological standpoint, but by asking a different question (especially one that inverts the way the statement or question is phrased) makes it possible to determine whether the subject is offering consistent responses that provide reliable insight from the same test, not going “outside of the test”, but to determine the estimates is by testing the Cronbach’s Coefficient Alpha ( a= 0.05), and Kuder-Richardson, and odd-even reliability . Because the scores often explains the differences of subvariables  of the given scores that are parallel to validate the reliability, so it is good to separate the correlation of the scores by categorize each items as odd and or even to test reliability of overall scores.

Content Validity

Content validity is one of the most important aspects of developing an assessment of a psychological construct such as passivity. In a multidimensional assessment of passivity, there are a variety of factors that just be considered. In that process, it is important to see the number of persons is in the same group of thought (negatively) and compare to the number of persons in the response of the given behavior (action) and then match both groups to the items reflected in the contents. There may be differences in how a subject perceives himself or herself and how that same subject responds to situations in real life. For example, a subject may respond to a question or statement in a way that indicates the subject sees himself or herself as passive. When provided with a question or statement about a particular scenario, however, such as being asked to agree or disagree with the statement that “if I found evidence my partner was cheating on me I would confront my partner with this evidence,” a response indicating agreement with this statement would conflict with the idea that the subject sees himself or herself as passive.

Because a multidimensional assessment of passivity is intended to examine the psychological construct from a verity of perspectives, it is important to have a broad and thorough understanding of passivity in order to develop an accurate assessment. The developer of the assessment must consider the many ways in which passivity is manifested in terms such as self-concept and actual behavior, as well as the different ways an individual might respond in the context of different types of relationships. For example, a subject who works as s supervisor might feel a sense of power and authority at work, and might be assertive and direct when dealing with subordinates. By contrast, this same subject might be in an abusive relationship wherein he or she feels powerless, and would be less likely to exhibit assertive behavior. The developer of an effective multidimensional assessment would take such factors into consideration to ensure that the assessment has adequate content validity.

Criterion Validity

In the most basic sense, the criteria of an assessment for the psychological construct of passivity should be related to real-world factors. If the response to the assessment indicates that the subject is psychologically passive, this should be correlative to scenarios in real-world situations. This means that there is some degree of predictive ability available when examining the results of the assessment, and that a subject with a high degree of passivity would be expected to respond to real-life situations in ways that demonstrate this passivity.  Which is the standard error of the estimate (SEest), predicting a failure to act and a tendency to challenge could increase the commonality behavior display in the society.

Construct Validity

The term “construct validity” refers to the relationship between the psychological construct being assessed (in this case it is passivity) and the way that the assessment is constructed. It is similar to, but not the same as, content validity and they can be seen as being interrelated. When making determinations about whether a subject is passive it is necessary to ensure that the assessment itself is appropriately constructed so that such determinations are valid. This means that testing must be grounded in theory, and that the results of the testing are measured against a theoretical construct

Convergent and Discriminant Validity

Convergent and Discriminant Validity are sub-types of Construct Validity. Simply put, they determine the degree to which measurements are related or unrelated to other measurements. In the context of Convergent Validity, measurements used to assess passivity, for example, should be weighed against other measurements of the same psychological construct across research literature. If they are closely correlated, this supports the Construct Validity of the assessment. By contrast, if a measurement of constructs that are supposed to be unrelated to each other can be seen as being unrelated to each other, that also supports the overall Construct Validity of the assessment or measurement. In measuring passivity, for example, the convergence of responses related to passivity will be higher in subjects with passivity. Responses to measurements of unrelated constructs should indicate that they are in fact unrelated. This Discriminant Validity also helps to ensure the overall Construct Validity of the assessment.

Item Analysis and Selection

Because there is some amount of subjectivity in making determinations and drawing inferences from a psychological assessment, it is important to be very careful about choosing the items that will be included on the assessment. Keeping all the factors in mind that have been previously discussed, selecting items for an assessment of passivity involves developing items that accurately target the issue being examined, and must also take into account the various factors that could skew the results. A subject who indicates a wide variety of different responses to questions about various scenarios may seem passive by one measure and aggressive by another. It is important to consider the statistical factors and other factors involved in order to accurately analyze the results and make determinations about the presence (or absence) of passivity.

Evaluating Gender, Ethnic, and Cultural Bias

While considering issues of gender, ethic, and cultural bias are important when developing any set of psychological measures or assessment, it is particularly important when measuring passivity.  For example, a psychological theory about passivity that is based on the responses of adolescent males may be an inappropriate basis for developing an item for a measurement of passivity to be administered to female subjects. Similarly, the individual conducting the assessment may have preconceived notions or expectations about how subjects from various groups (such as males/females, members of specific ethnic groups or cultural backgrounds, etc.) are likely to respond. Such biases can not only influence how responses are analyzed, they can also influence how assessments are developed in the first place.

The issue of potential biases shows how difficult it can be to measure a psychological construct such as passivity. It may be necessary, for example, to develop a variety of assessments that take various factors (gender, race, and others) into account, and to then compare the results of these various assessments to determine how to construct an assessment that is valid for a broad cross-section of subjects.

References

Brinkley, C., Schmitt, W., Smith, S., & Newman, J. (2001). Construct validation of a self-report psychopathy scale Personality And Individual Differences, 31.

Comm88.blogspot.com,. (2007). unidimensional vs. multidimensional variables. Retrieved 21 November 2014, from http://comm88.blogspot.com/2007/05/faq-unidimensional-vs-multidimensional.html

Homepages.rpi.edu,. (2014). Norming Distributions and Standardization. Retrieved 21 November 2014, from http://homepages.rpi.edu/~verwyc/Chap4btm.htm

Marsh, H., & Craven, R. (2006). Reciprocal Effects of Self-Concept and Performance From a Multidimensional Perspective. Beyond Seductive Pleasure and Unidimensional Perspectives. Perspectives On Psychological Science, 1(2), 133-163. doi:10.1111/j.1745-6916.2006.00010.x

Matlock-Hetzel, S. (2014). Basic Concepts in Item and Test Analysis. Ericae.net. Retrieved 21 November 2014, from http://ericae.net/ft/tamu/Espy.htm

Msu.edu,. (2014). Item Analysis. Retrieved 21 November 2014, from https://www.msu.edu/dept/soweb/itanhand.html

Psychology Wiki,. (2014). Test construction. Retrieved 21 November 2014, from http://psychology.wikia.com/wiki/Test_construction

Rusbult, C., &Zembrodt, I. (1983). Responses to dissatisfaction in romantic involvements: A multidimensional scaling analysis. Journal Of Experimental Social Psychology, 19(3), 274-293. doi:10.1016/0022-1031(83)90042-2