Common use of General Discussion Clause in Contracts

General Discussion. Except in rare instances of 100% participation, each peer nomina- tion study will require a decision regarding the treatment of non- participants. Whether to include or exclude nonparticipants as nominees may seem a minor methodological decision but has not been studied before. Historically, inclusion of nonparticipants has been the default decision and a fundamental requirement for valid peer nomination measures (Xxxxxxxxxxxxxx, 1943; Xxxxxx, 1934). Unfortunately, inclusion of nonparticipants raises ethical concerns (see Mayeux, Underwood, & Xxxxxx, 2007). We investigated two separate datasets, showing that the choice of including or excluding nonparticipants as nominees can affect psychometric properties of peer nominations. The effects varied between studies; differences between inclusion and exclusion were greater for certain types of missingness in Study 1 and greater in one school versus another in Study 2. Study 1 also indicated that, when the inclusion and exclusion conditions differed noticeably, exclusion consistently yielded lower reliability and different corre- lations than inclusion. From a psychometric perspective, our results indicate that non- participants should be included as potential nominees. Although exclusion of nonparticipants did not always detract from data qual- ity, it was likely to do so under conditions that are most probable in real-world situations; that is, when nonparticipants differ in status and peer preference from participants (Xxxx et al., 1997). Excluding nonparticipants as potential nominees removes the ability to test for systematic differences between participants and nonparticipants. Even if nonparticipants can be excluded as nominees without con- sequence when nonparticipation is completely random, the very exclusion of nonparticipants as nominees makes it impossible to demonstrate that nonparticipation is random. If our findings are representative of the research literature more generally, the fact that excluding low-status peers as nominees resulted in the greatest reduction in reliability and change in inter- correlations is concerning. Many peer relationships studies focus on youths who are rejected or unpopular, and they are least likely to participate in school-based research. Our findings highlight the importance of properly representing low-status or marginalized youths and, more generally, in maximizing participation rates with peer nominations. Our study used two large samples and investi- gated relatively low levels of missingness; however, because pre- vious studies have shown that higher levels of nonparticipation are associated with greater reductions in reliability (Marks et al., 2013) and internal validity (Xxxxxxx et al., 2018) of nomination measures, we expect that higher nonparticipation rates will result in even larger differences between including and excluding nonparticipants on rosters. Although the literature on practical solutions is sparse, Xxxxxx and Xxxxx (2017) recently suggested several strategies to deal with logistical hurdles (like low consent/participation rates) in peer nomination research. Limitations Although only one of our studies involved simulation of participant missingness, both involved simulating the exclusion of nonpartici- pants from rosters. When participants chose nominees, all peers were available on the roster; removal of choices happened post hoc. Nominations may have been different if nonparticipants were excluded from the rosters before data collection. While a key lim- itation, it is difficult to think of a methodologically sound way to test this issue without simulating roster exclusion. Additionally, there is currently no theoretical reason to believe that excluding nonparticipating peers from rosters would result in less error. The foundation of peer nominations is that nominators are comparing nominees to all other peers within a natural and closed social system (Cillessen & Marks, 2017; Xxxxxx, 1934). Providing participants with a non-random peer subgroup funda- mentally changes the nature of their choices. For example, it is difficult to predict how participants might react to naming peers they dislike when the most rejected peers are excluded from con- sideration. Some may choose fewer peers, which means losing valuable peer group data. Others may name the same number of less-disliked peers, in which case the sample distribution of rejec- tion nominations is biased. Regardless, the result is increased error and decreased accuracy. Another limitation of this investigation was that, due to at times double- and triple-layered non-statistical independence, it was not possible to conduct traditional significance tests of differences between inclusion and exclusion. Our conclusions are based on bootstrapped distributions comparing randomly missing nominees to systematically missing nominees, which is not quite the same as comparing systematically missing nominees to no missing xxxx- xxxx. Future research might explore different data structures or new statistical tests to account for the multiple layers of dependency of our current study. Research Ethics This study focused on the psychometric impact of including versus excluding nonparticipants as nominees in peer nomination research; however, there is also an ethical impact to collecting data about individuals not participating in a study. We were fortunate to have two datasets collected using passive consent procedures. Only a small handful of students failed to participate because their parents actively withheld consent, and we acknowledge that these parents may not have realized that secondary data were being collected on their children anyway. Most “nonparticipants” who served as xxxx- xxxx, however, were either absent for data collection or did not provide any nominations. In an active consent procedure, the partic- ipation rates would be lower and a larger proportion of participant missingness would have been due to lack of parental consent. Treating individuals without consent to participate in a study (particularly those who have been actively denied consent) as xxxx- xxxx in peer nomination measures results in identifiable data col- lection for them. This could be a violation of a key principle of informed consent in behavioral research. However, the violation of this principle does not, in-and-of-itself, cause a research methodol- ogy to be unethical. Previous research has indicated that the risks of being involved in sociometric research are no greater than risks faced by children and adolescents in everyday life (Xxxxxx et al., 2007), and the benefits of peer nomination research are substantial. Peer measurements provide a unique perspective on social beha- viors among adolescents (particularly behaviors hidden from teach- ers or observers, such as relational aggression) and are irreplaceable to assess affective and status variables like friendship, social pre- xxxxxxx, and popularity (Xxxxxxxxx & Xxxxx, 2017). Moreover, sociometric data can play an important role in solving school prob- lems (bullying prevention, identifying at-risk students, etc.). The possible deviation from principles of informed consent is a cost that cannot be ignored. This cost, however, should be weighed against the benefits of collecting data on social status, relationships, and behaviors that other methods cannot measure. If strictly main- taining informed consent (i.e. excluding nonparticipants as xxxx- xxxx) undermines the quality of peer nomination data, then invalid measurements will be the cost of avoiding the ethical conflict. Ultimately, we recommend that researchers take a holistic approach to the cost/benefit analysis inherent in peer nomination research and consider psychometric and theoretical concerns alongside ethical ones. Our investigation presented information relevant to this cost/benefit analysis by quantitatively assessing the differences between including and excluding nonparticipants as nominees. Two studies showed that data quality can be negatively affected by the choice of whether nonparticipants are included as nominees. We hope that this research will spark further discussion and investigation of peer nomination methodology and of the inter- section between methodology and ethics in social developmental research.