J Happiness Stud (2008) 9:219–226 DOI 10.1007/s10902-007-9047-4 RESEARCH PAPER What wealth-happiness paradox? A short note on the American case Claude S. Fischer Published online: 20 February 2007 Ó Springer Science+Business Media B.V. 2007 Abstract Happiness scholars have tried to resolve the seeming paradox that as Americans’ wealth increased substantially over the last few decades, their happiness did not. This article questions whether the paradox is real. Demonstrations of the paradox almost always rely on GDP per capita as the measure of wealth, but that is a poor measure of a people’s well-being. It is heavily and increasingly skewed; it does not account for effort. Using instead measures of household income, male income, and average wages eliminates the paradox; these indicators of affluence have grown only slowly or declined in the same period, paralleling the changes in happiness scores. Moreover, using these indicators reveals a modest but real correlation between material well-being and national happiness. Keywords Happiness Æ Income Æ Paradox Æ Easterlin Æ Wealth Æ Measurement Æ Method Many scholars have been trying to resolve the wealth-happiness paradox first identified in the 1970s by Easterlin (1973) and replicated ever since: Over the last few decades, it seems, Americans’ wealth increased substantially, but their happiness, as measured in surveys, did not.1 (The paradox has been identified for other nations as well, but I restrict myself here to the United States.) Researchers’ answers to the paradox include psychological explanations for why greater income, at least beyond some threshold, fails to make people happier; for example, people’s expectations for affluence may rise because of adaptation or because of social comparisons so as to offset advances in actual affluence. Another category of 1 See also Easterlin (2001), Lane (2000), Layard (2005), Frey and Stutzer (2002), DiTella, MacCulloch, and Oswald (2003), Binswager (2006), Hagerty (2000), Hagerty and Veenhoven (2003), Veenhoven (2000, 2005), Schwartz (2004); Ott (2001); Alesina, Di Tella, and MacCulloch (2001), Oswald (1997), etc. C. S. Fischer (&) Department of Sociology, University of California, Berkeley, CA, USA e-mail: fischer1@berkeley.edu 123 220 C. S. Fischer answers posits that rising wealth did make Americans happier, but contemporaneous and depressing changes, such as increasing divorce rates or declining sociability, canceled out the euphoria of greater affluence. I argue here, in contrast, that the there is no paradox to be explained; it is an illusion based on mis-specifying material well-being. Repeatedly, the paradox appears when the time trend for happiness is juxtaposed to the time trend for GDP per capita. GDP per capita, however, is an inappropriate measure of people’s material well-being. First, using GDP per capita ignores the skewed distribution of the domestic product and its increasing skewness over time. Second, using GDP per capita ignores the cost in effort, the personal investment, required to gain the wealth. Once these are problems are addressed, the paradox evaporates. I present below a simple demonstration of this argument, leaving aside the subtleties of measurement and modeling.2 The basic point is clear enough. (For a more sophisticated treatment of some of these concerns, see Hout (2006).) The sources of the data I use are described in the note to this sentence.3 Happiness and income, 1972–2005 Figure 1 is a version, carried through 2005, of the standard display that greater wealth did not bring greater happiness. The happiness measure, the one most commonly used, is the General Social Survey’s ‘‘HAPPY’’ item, asked about two dozen times over 32 years: ‘‘Taken all together, how would you say things are these days—would you say that you are very happy, pretty happy, or not too happy?,’’ re-scored from 1, ‘‘not too,’’ to 3, ‘‘very happy’’ (right-hand scale). The wealth measure is GDP per capita (in ‘‘chained’’ 2000 dollars; left-hand scale). The lines represent simple, linear regressions to summarize the trend. Here is the seeming paradox: GDP per capita, adjusted for inflation, grew greatly; the 2003–05 average was 1.8 times the 1972–74 average. Average national happiness grew not at all in the same period; it may even have declined. But GDP per capita is, I contend, a woeful representation of the general population’s material well-being. A major reason is that income is heavily skewed. In 2005, the top 20% of income recipients took in 50% of the national income; the person at the 95th percentile of income brought home 3.61 times what the person at the 50th percentile did. And that skew had increased since 1972, when the estimates were 44% of national income and a 95:50 ratio of 2.75 (DeNavas-Walt et al., 2006, Table A-3). Moreover, wealth—i.e., 2 I refer to issues such as measurement anomalies in the GSS happiness index (Smith, 1979), preserving the categorical quality of the measure and, applying time-series appropriate models. 3 Happiness, GSS: The mean score for the national sample of the General Social Survey’s HAPPY question, 1972–2004. Happiness, 1946–2005 (for Figure 5): Means for non-GSS surveys from the World Database of Happiness (Veenhoven, 2006). GDP per capita: Gross Domestic Product in chained 2000 dollars, from the Bureau of Economic Analysis (http://www.bea.gov/bea/dn/home/ gdp.htm), divided by annual population. Median household income: DeNavas, Proctor, and Lee (2006), table A1. Median male income: Census historical tables, http://www.census.gov/hhes/www/ income/histinc/p02.html, table P2. Mean hourly wages, 1947–99: Historical Statistics of the United States, Millennial Edition, Online (http://www.hsus.cambridge.org/HSUSWeb/toc/hsusHome.do), table C Ba4440-4483—Hourly and weekly earnings of production workers in manufacturing, by industry: 1947B1999. Mean hourly wages, 2000–05: Bureau of Labor Statistics, Employment, hours, and earnings from the current employment statistics (http://www.data.bls.gov/PDQ). I calculated inflation adjustments, where needed, using the CPI-U series. 123 What wealth-happiness paradox? 221 2.4 $50,000 2.35 $40,000 2.25 $30,000 2.2 $20,000 2.15 M ean GSS Happiness Score GDP per capita (2000 dollars) 2.3 2.1 $10,000 GDP/cap 2.05 Happy mn $0 1970 1975 1980 1985 1990 1995 2000 2005 2 2010 Fig. 1 Mean happiness and GDP per capita, 1972–2005 assets minus debts—is skewed much more than income is and it became increasingly skewed after 1972 (see, e.g., Fischer & Hout, 2006, Ch 6). Why would we expect that growing wealth going to a smaller and smaller proportion of the population would raise average happiness? For these—and other4—reasons, GDP per capita is poor way to assess a people’s material well-being. A rough way to get around this problem is to replace GDP per capita with median household income. Figure 2 does that, using 2005 dollars. The affluence-happiness paradox remains, but it is no longer a stark contrast; median household income in 2003–05 was only 1.2 times that of 1972–74. Also, we see more variability from year to year in household income than we observe in GDP per capita, variability which might be related to variability in happiness (more on this later). The next step is to include in our considerations what we know about how American households kept their incomes up in this era. To a great extent, they did it by adding workers and hours. Reports from the field suggest that sluggishness in male breadwinners’ incomes was an impetus to wives working—and also a barrier to marriage in many cases (e.g., Jacobs & Gerson, 2001). Average Americans put more painful effort (e.g., in commuting, off-hour shifts, child-care arrangements) into making money than they did before the 1970s. Such effort and pain should be factored in on the negative side of the income ledger to really measure net material well-being. One simple way to index that effort is to use, instead of median household income, median male incomes. That measure indirectly corrects for the growing female work contribution. Figure 3 displays the result. The paradox becomes yet less paradoxical. Median male income grew only slightly; the 2003–2005 average was 1.1 times the 1972–1974 average. And we see, again, the cycles that are hardly visible in the GDP data. Alternatively, we can use 4 There are other reasons to reject GDP per capita for such analyses: The happiness surveys are of adults, but GDP per capita includes children in the denominator. Relatedly, the GDP per capita figure rises just because the birth rate drops. Also, an increasing proportion of GDP was held by corporations in a form not easily accessible even to shareholders. 123 222 C. S. Fischer $50,000 2.4 2.3 2.25 $30,000 2.2 $20,000 2.15 2.1 Mean GSS Happiness Score Median Household Income (2005 dollars) 2.35 $40,000 $10,000 med hh inc 2.05 Happy mn $0 1970 1975 1980 1985 1990 1995 2000 2005 2 2010 Fig. 2 Median household income and happiness, 1972–2005 2.4 $50,000 2.35 2.3 2.25 $30,000 2.2 $20,000 2.15 Mean GSS Hap pin ess Sco re Med ian Male Inco me (2004 d o llars) $40,000 2.1 $10,000 male income 2.05 Happy mn $0 1970 1975 1980 1985 1990 1995 2000 2005 2 2010 Fig. 3 Median male income and happiness, 1972–2004 for similar purposes median male earnings for full-time workers. That pattern (not shown) is, as we would expect, much like the one in Fig. 3. The growth in inflationadjusted median male earnings from 1972 to 2005 was all of 1 percent. Another take would be to go gender-neutral and look at overall hourly wages for both men and women as a crude index of the income/effort ratio. Average hourly earnings, inflation-adjusted, of all production or non-supervisory workers in private industry dropped by one-tenth from 1972–74 to 2003–05; it dropped more than mean happiness. These data are displayed in Fig. 4. In sum, the closer one approximates 123 What wealth-happiness paradox? 223 $25.00 2.4 2.3 2.25 $15.00 2.2 $10.00 2.15 2.1 Mean GSS Hap pin ess Sco re Mean Ho urly Wag es (2000 d o llars) 2.35 $20.00 $5.00 wage series 2.05 Happy mn $0.00 1970 1975 1980 1985 1990 1995 2000 2005 2 2010 Fig. 4 Mean hourly wages and happiness, 1972–2005 changes in the real, net material well-being of average Americans, the closer the trend line for affluence looks like that of happiness—and the paradox evaporates. Another rough-and-ready approach is to simply correlate, over time, mean happiness with the various measures of material well-being. The following table does so, in a few different ways. The first column displays the correlations over the years 1972–2004 between mean happiness in the GSS and the various income measures I have discussed. The top two entries of the first column shows that there is negligible correlation between annual happiness and either annual GDP per capita or annual median household income over a period when both the income measures grew —another display of the famous ‘‘paradox.’’ The bottom three rows of the column, however, show a modest but real positive correlation of happiness with the other—and, I argue, better—indicators of material well-being. Inspection of the scatter plots in Figs. 1–4 suggests that there are two notable outliers in the happiness data. One is 2002: the GSS asked people how happy they were about six months after 9/11 and the depressive effects of the event seem clear. Another outlier is 1972: it is one of the years in which the ‘‘happy’’ question was in an unusual sequence and the General Social Survey recommends it be dropped from trend analyses (Smith, 1979). The 1972 point also has special statistical leverage because it is first in the series. The last three columns of Table 1 display the correlations holding out those points. They suggest that GDP per capita was actually a negative predictor of happiness—if we exclude 1972—and that median household Table 1 Correlations, 1972–2004, of mean happiness with various economic measures GDP per capita Median house income Median male income Median male earnings Mean hourly wage 1972–2004 Excl. 2002 Excl. 1972 Excl. 1972 & 2002 –.07 .03 .21 .20 .19 –.01 .09 .29 .26 .19 –.22 –.06 .30 .19 .45 –.16 .00 .40 .26 .46 123 C. S. Fischer $40,000 2.8 $30,000 2.6 $20,000 2.4 $10,000 $0 1945 M ean Happiness Score Constant Dollars 224 2.2 1955 1965 GDP/cap male income hourly wages Happy mn 1975 1985 1995 2 2005 Fig. 5 GDP per capita, median male income, hourly wages (annualized), and happiness, c. 1947–2005 income was unrelated to happiness. Notably, these columns also show that the other three measures of economic well-being positively predicted the national rate of happiness. Exploring a longer series Veenhoven’s World Database of Happiness provides a longer series of happiness scores from the same question, although they are from a disparate set of sources and much less consistent than the GSS. Nonetheless, they allow us to extend the window from about 30 to almost 60 years. I have taken the average happiness score for each year available, excluding those from NORC or GSS surveys so as not to overlap with the prior analysis and not to overweigh the last three decades. The happiness scores are displayed, from 1946 to 2005, in Fig. 5 as open circles. (This series is far noisier than the GSS series for 1972–2004.) The best-fitting smoother is a cubic function.5 I used and display three income series that are about the same length, also applying cubic smoothers for consistency.6 We see, first, that the GDP per capita trend (black boxes) does not parallel the happiness ‘‘trend’’; the correlation is –.08 (.02 if we drop the poll taken in November, 2001, two months after 9/11). However, median male income (black diamonds) is a closer fit (r = .11; .23 without November, 2001), as is average hourly wages for production workers in manufacturing-multiplied by 2000 hours for better presentation (black circles; r = .11; .13 without 2001). More broadly, the data hint that happiness was increasing during the years of the postwar boom when affluence was also becoming more broadly shared and then happiness leveled off as affluence did for the mainstream population and inequality grew. But the happiness measures are too scattered and rough to make this more than a tentative suggestion. 5 The R-squared for mean happiness score X year is .03 for a linear regression, .05 for a quadratic one, and .16 for a cubic one (.20 if the November, 2001, poll is dropped). 6 The ‘‘hourly wages’’ series here is for manufacturing only, because of data limitations. 123 What wealth-happiness paradox? 225 Conclusion This back-of-the-envelope exercise provides only a crude picture of the happiness– affluence connection in late twentieth- and early twenty-first-century America. Fuller explorations would require finer measures and more complex time-series and multivariate analyses. Still, it seems clear that the ‘‘paradox’’ which has perplexed so many is not such a paradox after all. GDP growth has been expansive and sustained, but GDP has become increasingly unevenly distributed. At the same time, the task of keeping up with living standards and coping with sagging male earnings has required much more strenuous efforts by average American families. When we take these points into even only approximate consideration, we can see some evidence that national happiness stalled because the income/effort balance stalled for average families. We can also see that, while many things depress or raise Americans’ reports of their happiness (e.g., 9/11; adaptation; social comparisons), fluctuations in material well-being is one them. Readers of an earlier draft have wisely pointed out limitations of this analysis. For example, a wealth-happiness paradox appears to exist in some, although not most, other countries. I leave those cases to others. Also, my analyses here are only preliminary; more sophisticated treatments are called for. Nonetheless, these data serve, I hope, to demonstrate that researchers must define and measure material well-being much more accurately and that the ‘‘fact’’ which has driven so much theoretical diagnoses, the supposed wealth-happiness ‘‘paradox’’ in America, is yet to be demonstrated to a be a fact. Acknowledgment I appreciate comments on an earlier draft by Richard Easterlin, Michael Hout, Ruut Veenhoven, and Rafael Di Tella, but I, of course, remain solely responsible for errors of understanding and method. References Alesina, A., Di Tella, R., & MacCulloch, R. (2001). Inequality and happiness: Are Europeans and Americans different? National Bureau of Economic Research Working Paper No. W8198 Binswanger, M. (2006). Why does income growth fail to make us happier? Searching for the treadmills behind the paradox of happiness. Journal of Socio-Economics, 35, 366–381 DeNavas-Walt, C., Proctor, B. D., & Lee, C. H. (2006). Income, poverty, and health insurance coverage in the United States: 2005. U.S. Census Bureau, Current Population Reports, P60-231, U.S. Government Printing Office, Washington, DC Di Tella, R., MacCulloch, R. J., & Oswald, A. J. (2003). The macroeconomics of happiness. Review of Economics and Statistics, 85, 809–827 Easterlin, R. A. (1973). Does money buy happiness? The Public Interest, 30, 3–10 Easterlin, R. A. (2001). Income and happiness: toward a unified theory. The Economic Journal, 111, 465–484 Fischer, C. S., & Hout, M. (2006). Century of difference: How America changed in the last hundred years. New York: Russell Sage Foundation Frey, B. S., & Stutzer, A. (2002). Happiness and economics. Princeton, NJ: Princeton University Press Hagerty, M. (2000). Social comparisons of income in one’s community: Evidence from national surveys of income and happiness. Journal of Personality and Social Psychology, 78, 764–771 Hagerty, M., & Veenhoven, R. (2003). Wealth and happiness revisited—growing national income does go with greater happiness. Social Indicators Research, 64, 1–27 123 226 C. S. Fischer Hout, M. (2006). Money and morale. Working Paper, Survey Research Center, University of California, Berkeley Jacobs, J. A., & Gerson, K. (2001). Overworked individuals or overworked families? Explaining trends in work, leisure, and family time. Work and Occupations, 28, 40–63 Lane, R. E. (2000). The loss of happiness in market democracies. New Haven: Yale University Press Layard, R. (2005). Happiness: Lessons from a new science. New York: Penguin Oswald, A. J. (1997). Happiness and economic performance. Economic Journal, 107, 1815–1831 Ott, J. (2001). Did the market depress happiness in the US? Journal of Happiness Studies, 2, 433–443 Schwartz, B. (2004). The paradox of choice: Why more is less. New York: Ecco Smith, T. W. (1979). Happiness: Time trends, seasonal variations, intersurvey differences, and other mysteries. Social Psychology Quarterly, 42, 18–30 Veenhoven, R. (2000). Freedom and happiness: A comparative study in 46 nations in the early 1990’s. In E. Diener & E. M. Suh (Eds.), Culture and subjective well-being (pp. 257–288). Cambridge, MA: MIT Press Veenhoven, R. (2005). Is life getting better?: How long and happily do people live in modern society? European Psychologist, 10, 330–343 Veenhoven, R. (2006). World database of happiness. Erasmus University Rotterdam. Available at: http://worlddatabaseofhappiness.eur.nl 123 1 Understanding Research Assignment #2 Correlational Research Instructions: Reading research for the first time can be difficult. Even as a professor who has read a lot of research over the years, I often have to read the article more than once for it all to click. So don't be concerned that you are having to read slowly, carefully, and more than once -- that's normal. Psychology is a science based on research and this class is about the methods scientists use to conduct research, so it's good for students to read actual studies. Read the study and attempt to answer the questions on your own first. Please write your responses in blue font so that it is easier for me to grade. A. Variables 1. Identify the independent variable and the dependent variable for an experimental study. For a non-experimental study, what were the variables of interest? (What were the levels of the independent variable – there has to be at least two. The IV = the variable that differs between the groups. It is often the treatment. The DV = what the experimenters used to determine if the IV had an effect. These are often formal assessments.) a. IV = b. DV = OR for non-experimental research designs c. The variables of interest = 2. Summarize the statement of the problem. (This is the researcher’s justification of the why the experiment is needed. Typically, this will be in the Introduction.) 3. Identify the research question (The hypothesis. This should be a question. What did the experimenter’s want to know? The title of the article can be a shortened version of the research question. Typically, the research question is in the Introduction.) B. Method Section 1. Summarize the Method Section. (Read the entire Method Section. You should describe what happened to the participants at the start of the study, the treatment, and the end of the study.) 2 C. Research Design 1. Identify the research design. (First, identify the overall category. Was this a descriptive study? A correlational study? An experimental study? A quasi-experimental study? A single-subject design?) Once you have identified the overall category, you need to specify the exact type. For example, “This was a descriptive study and used naturalistic observation.” a. Type of research = b. Specific research design = D. Results/Discussion 1. How were the results recorded? Did the authors use a table or figure? If so, what was in it? What formal tests did they use to measure the variables? For example, “The researchers used the Vinland Adaptive Behavior Scales to measure the participants overall level of functioning in daily life and the pre and post scores were displayed in a table.) If applicable, what statistical tests were used? (Was it a t-test? An ANOVA?) 2. Summarize the results. (Read the entire Results section. What happened?) 3. What is a conclusion you can take from the study? (What is the take-home message from this study? How can the findings be used by non-researchers?) 4. Based on reading this study, what is a needed, future research topic? (Hint: many authors towards the end of the Discussion section talk about their view of needed future research. You can use one of theirs or a research question that is your own.) JOURNAL OF APPLIED BEHAVIOR ANALYSIS 2017, 50, 176–180 NUMBER 1 (WINTER) EFFECTS OF THE GOOD BEHAVIOR GAME ACROSS CLASSROOM CONTEXTS BRITTANY PENNINGTON AND JENNIFER J. MCCOMAS UNIVERSITY OF MINNESOTA The Good Behavior Game (GBG), a well-researched classroom group contingency, is typically played for brief periods of time, which raises questions about the effects on subsequent contexts. This study used a multiple baseline design and showed that when the GBG was implemented in one context, behavior improved in only that context. Behavior improved in the subsequent activity only when the GBG was implemented. Key words: classroom behavior, general education, interdependent group contingency, ontask behavior, schools The Good Behavior Game (GBG) is an interdependent group contingency in which the teacher divides the class into teams, establishes rules for behavior, and distributes points either for appropriate behavior (Fishbein & Wasik, 1981) or for problem behavior (Barrish, Saunders, & Wolf, 1969). At the end of the game, the teacher delivers prizes to members of the team with the most or fewest points, depending on the version of the game (Tanol, Johnson, McComas, & Cote, 2010). The GBG is typically implemented for a brief period of time during a particular activity or setting, such as 10 to 35 min during whole group lessons (e.g. Donaldson, Vollmer, Krous, Downs, & Berard, 2011; Tanol et al., 2010), 10 to 15 min in the cafeteria (McCurdy, Lannie, & Barnabas, 2009), or approximately 60 min during math class (Flower, McKenna, Muething, Bryant, & Bryant, 2014). Although The research described in this article was supported in part by Grant H325H140001 from the Office of Special Education Programs, U.S. Department of Education. Nothing in the article necessarily reflects the positions or policies of the federal government, and no official endorsement by it should be inferred. Address correspondence to: Brittany Pennington, Department of Educational Psychology, 250 Education Sciences Building, 56 East River Road, Minneapolis, MN, 55455 (e-mail: penni156@umn.edu). doi: 10.1002/jaba.357 researchers have demonstrated that the GBG effectively reduces problem behavior or increases appropriate behavior when it is in effect (Flower, McKenna, Bunuan, Muething, & Vega, 2014), the fact that it is implemented only for a brief period of time might raise some issues. For example, it is plausible that behavior could worsen following the game due to decreased availability of reinforcement for appropriate behaviors or increased availability of reinforcement for problem behaviors. By contrast, it is plausible that the game’s effect could continue after the teacher has stopped playing the game. Additionally, research has not shown whether the game could maintain appropriate behavior for individual students across activities in a classroom. Donaldson, Wiskow, and Soto (2015) examined the effect of the GBG on disruptive behavior immediately before and after the GBG was implemented. They found that disruptive behavior was suppressed during the game but was unchanged in the time immediately before or after the game was played. However, the authors collected class-level data and recommended further research on individual students’ behavior following the GBG. The current study extends previous research on the GBG by using a multiple baseline design to examine the effects of the GBG on three students’ on-task behavior across subsequent contexts in a classroom. © 2016 Society for the Experimental Analysis of Behavior 176 GOOD BEHAVIOR GAME ACROSS CONTEXTS METHOD Participants, Setting, and Materials Although all students in the classroom played the game, we collected data on the behavior of three third-grade students whom the teacher reported were the most off-task students in the class. All three participants, Aquila, Ishkode, and Debwewin, were 8 years old and of Native American descent. Aquila (female) was diagnosed with an emotional behavioral disorder (EBD). Ishkode and Debwewin (both male) were identified by the school’s problem solving team as at risk for being diagnosed with EBD. The teacher (female) had 4 years of teaching experience. The aide (male) had been a classroom aide for 1 year. Due to some students moving during the school year, between 15 and 18 students were enrolled in the class during the study. The study was conducted in a public school in a large, urban district with 94% of the students eligible for free or reduced lunch. We conducted this study in the classroom during morning meetings and math rotations. During morning meetings, the students sat in a circle on the carpet while the teacher led the activities. Morning meeting lasted 10 to 20 min. During each meeting, students greeted one another, discussed the schedule for the day, solved math problems, and played a brief game. There was no seating assignment during morning meetings, but beginning in session 9, students sat with their team, with one team comprising one half of the circle, and the other team comprising the other half of the circle. During math rotations, the students worked in small groups: One group worked with the teacher, one group worked with the aide, and one group played math games with partners or on iPad® tablets. After 20 min, the groups rotated. During the math rotation observations, Aquila worked with the aide while Ishkode and Debwewin played math games with partners or on iPad® tablets. The teacher, aide, and first author implemented the game collaboratively: All three 177 watched for on-task behavior and marked points on small white boards with dry erase markers. Prizes included small pieces of candy, pencils, and figurines. Dependent Variable During morning meetings, data collectors recorded on-task behavior if students were sitting on the floor and looking at the teacher or speaker, making relevant on-topic comments, following the rules described by the teacher at the beginning of a community-building game, or making encouraging comments to classmates. During math rotations, data collectors recorded on-task behavior if students were at the assigned location and oriented towards the group with the appropriate materials, used the materials appropriately, or made encouraging comments. Data Collection On-task data were collected using 10-s momentary time-sampling procedures, rotating among the three students (i.e., every student was observed every 30 s). Researchers used a stopwatch or iPhone and earpiece to deliver a quiet beep at each 10-s interval. Data were collected for the first 10 min of morning meetings and for the first 10 min of the first math rotation. However, when we implemented the GBG, the game was played for the duration of the morning meeting and for all three math rotations. Sessions were conducted 3 days per week for 7 weeks. Design and Procedures We used a multiple baseline design to analyze the effects of the GBG on on-task behavior during morning meetings and then during math rotations. During the GBG, all the students in the classroom were divided into two teams and student names were written under their teams’ names on a poster in the front of 178 BRITTANY PENNINGTON and JENNIFER J. MCCOMAS the room. Two participants were on one team and the third participant was on the other team; team membership did not change during the study. Baseline During baseline, the teacher and the aide reminded students of the classroom rules but did not deliver points or offer prizes. The teacher and the aide either verbally corrected or ignored off-task behavior. Intervention First, we introduced the GBG during morning meetings. The teacher divided the class into two teams, with eight to nine students on each team. A poster in the front of the classroom listed the behavior expectations, which were aligned with the Native American grandfather teachings and had been explicitly taught and periodically reviewed (e.g., respect, wisdom, and love), and which students were on each team. The teacher, aide, and first author each carried a small white board and a dry erase marker with a line down the middle and the team names on top. The teacher explained that when a whole team was on-task, the team could earn a point, and the team with the most points at the end of the morning could earn a prize to be awarded approximately 2 hr after math as students departed for lunch. The teacher, aide, or first author awarded points approximately every 30 s to teams that had all team members on task. The first author instructed the teacher and aide to deliver points about every 30 s to teams that were on task, that they could take turns delivering points, or that the first author could deliver points. The first author also instructed the teacher and the aide that they should not both give points at the same time. The teacher, aide, and first author communicated to determine who would award points; the first author delivered points most frequently. The first author had a timer that vibrated every 10 s, and she either awarded points or cued the teacher or aide to award points every third vibration. If 1 min passed with no points awarded and one or both teams were on-task, the first author started awarding points and continued until the teacher or the aide was ready to resume awarding points. At the end of the game, the first author added the points from all three white boards to determine the winning team. Interobserver Agreement and Fidelity Interobserver agreement (IOA) data were collected for 42% of morning meeting sessions and 29% of math rotation sessions. IOA for on-task behavior was calculated by dividing the number of intervals during which the observers agreed by the total number of intervals, and multiplying by 100. IOA averaged 85% during morning meetings (range 78%-94%) and 88% during math rotations (range 83%-94%). Trained observers other than the first author collected procedural fidelity data using a sixitem checklist, during 33% of sessions; average procedural integrity was 97% (range 83%-100%). The checklist included telling students the game was beginning, only awarding points during the specified game time, awarding points when the whole team or all team members within a group were on task, not awarding points when team members were off task, informing students when the game was finished, and delivering prizes to the winning team. RESULTS AND DISCUSSION During morning meetings (Fig. 1), baseline data for on-task behavior were variable. Aquila was on-task in 83% of intervals (range, 10%100%), Debwewin was on-task in 40% of intervals (range, 27%-64%), and Iskhode was on-task in 55% of intervals (range, 17%-83%). When we implemented the GBG, on-task behavior increased and variability decreased for all three participants. On average, Aquila was on-task in GOOD BEHAVIOR GAME ACROSS CONTEXTS GBG BL GBG BL GBG BL 179 100 80 Morning Meeting On-task Behavior (% of intervals) 60 Aquila 40 Ishkode Debwewin 20 0 100 80 60 Math 40 20 0 -20 1 3 5 7 9 11 13 15 17 19 21 1 !"#$ 3 5 7 Sessions 9 11 13 15 17 19 21 Sessions !"#$ 1 3 5 7 9 11 13 15 17 19 21 Sessions Figure 1. Percentage of intervals with on-task behavior during morning meeting and math rotations during baseline (BL) and the Good Behavior Game (GBG). 99% of intervals (range, 93%-100%), Debwewin was on-task in 80% of intervals (range, 32%-100%), and Ishkode was on-task during 89% of intervals (range, 72%-100%). No change in baseline on-task behavior was apparent for any of the participants during math rotations when the GBG was implemented during the immediately preceding morning meeting. During math rotation baseline, on average, Aquila was on-task in 60% of intervals (range, 29%-100%), Debwewin was on-task in 33% of intervals (range, 0%-72%), and Ishkode was ontask in 56% of intervals (range, 17%-76%). When the GBG was implemented during math rotations, the level of on-task behavior increased and variability decreased for all three participants. On average, Aquila was on-task in 96% of intervals (range, 83%-100%), Debwewin was on-task in 91% of intervals (range, 84%-100%), and Ishkode was on-task in 88% of intervals (range 16%-100%). Missing data points indicate the student was absent during that session. Consistent with previous research, the findings of this study suggest that teachers who implement the GBG can expect increased on-task behavior while the game is in effect (Flower, McKenna, Bunuan, et al., 2014). Donaldson et al. (2015) showed that class-level disruptive behavior neither increased nor decreased before or after the GBG. This study extends that finding by suggesting that on-task behavior of individual students is not likely to improve or to worsen in activities following the GBG. Additionally, this study demonstrates that the effects of the GBG can be observed when implemented across multiple contexts in a classroom. Several limitations should be noted. First, we collected data on the behavior of only three students in one class; different students may respond differently following the GBG. Second, it is unclear what contingencies maintained off-task behavior during baseline; this information may enhance the design of the GBG. 180 BRITTANY PENNINGTON and JENNIFER J. MCCOMAS Third, responding following the GBG might have been different had the subsequent activity been more similar to the activity during which the GBG was implemented. Future researchers could evaluate different ways to program for maintenance and generalization of children’s appropriate behavior following the GBG and in different contexts. Programming common stimuli during GBG and non-GBG times or using indiscriminable contingencies could facilitate a spread of the effects of the GBG to non-GBG times or to other contexts in school. REFERENCES Barrish, H. H., Saunders, M., & Wolf, M. M. (1969). Good Behavior Game: Effects of individual contingencies for group consequences on disruptive behavior in a classroom. Journal of Applied Behavior Analysis, 2, 119-124. http://doi.org/10.1901/jaba. 1969.2-119 Donaldson, J. M., Vollmer, T. R., Krous, T., Downs, S., & Berard, K. P. (2011). An evaluation of the Good Behavior Game in kindergarten classrooms. Journal of Applied Behavior Analysis, 44, 605-609. http://doi.org/10.1901/jaba.2011.44-605 Donaldson, J. M., Wiskow, K. M., & Soto, P. L. (2015). Immediate and distal effects of the Good Behavior Game. Journal of Applied Behavior Analysis, 48, 685689. http://doi.org/10.1002/jaba.229 Fishbein, J. E., & Wasik, B. A. (1981). Effect of the Good Behavior Game on disruptive library behavior. Journal of Applied Behavior Analysis, 14, 89-93. http://doi.org/10.1901/jaba.1981.14-89 Flower, A., McKenna, J. W., Bunuan, R. L., Muething, C. S., & Vega, R. (2014). Effects of the Good Behavior Game on challenging behaviors in school settings. Review of Educational Research, 84, 546-571. http://doi.org/10.3102/003465431453 6781 Flower, A., McKenna, J., Muething, C. S., Bryant, D. P., & Bryant, B. R. (2014). Effects of the Good Behavior Game on classwide off-task behavior in a high school basic algebra resource classroom. Behavior Modification, 38, 45-68. http://doi.org/10.1177/ 0145445513507574 McCurdy, B. L., Lannie, A. L., & Barnabas, E. (2009). Reducing disruptive behavior in an urban school cafeteria: An extension of the Good Behavior Game. Journal of School Psychology, 47, 39-54. http://doi.org/ 10.1016/j.jsp.2008.09.003 Tanol, G., Johnson, L., McComas, J., & Cote, E. (2010). Responding to rule violations or rule following: A comparison of two versions of the Good Behavior Game with kindergarten students. Journal of School Psychology, 48, 337-355. http://doi.org/10.1016/j.jsp. 2010.06.001 Received March 31, 2016 Final acceptance September 13, 2016 Action Editor, Jeanne Donaldson 1 Understanding Research Assignment #3 Single Subject Design – Multiple Baseline Instructions: Reading research for the first time can be difficult. Even as a professor who has read a lot of research over the years, I often have to read the article more than once for it all to click. So don't be concerned that you are having to read slowly, carefully, and more than once -- that's normal. Psychology is a science based on research and this class is about the methods scientists use to conduct research, so it's good for students to read actual studies. Read the study and attempt to answer the questions on your own first. Please write your responses in blue font so that it is easier for me to grade. A. Variables 1. Identify the independent variable and the dependent variable for an experimental study. For a non-experimental study, what were the variables of interest? (What were the levels of the independent variable – there has to be at least two. The IV = the variable that differs between the groups. It is often the treatment. The DV = what the experimenters used to determine if the IV had an effect. These are often formal assessments.) a. IV = b. DV = OR for non-experimental research designs c. The variables of interest = 2. Summarize the statement of the problem. (This is the researcher’s justification of the why the experiment is needed. Typically, this will be in the Introduction.) 3. Identify the research question (The hypothesis. This should be a question. What did the experimenter’s want to know? The title of the article can be a shortened version of the research question. Typically, the research question is in the Introduction.) B. Method Section 1. Summarize the Method Section. (Read the entire Method Section. You should describe what happened to the participants at the start of the study, the treatment, and the end of the study.) 2 C. Research Design 1. Identify the research design. (First, identify the overall category. Was this a descriptive study? A correlational study? An experimental study? A quasi-experimental study? A single-subject design?) Once you have identified the overall category, you need to specify the exact type. For example, “This was a descriptive study and used naturalistic observation.” a. Type of research = b. Specific research design = D. Results/Discussion 1. How were the results recorded? Did the authors use a table or figure? If so, what was in it? What formal tests did they use to measure the variables? For example, “The researchers used the Vinland Adaptive Behavior Scales to measure the participants overall level of functioning in daily life and the pre and post scores were displayed in a table.) If applicable, what statistical tests were used? (Was it a t-test? An ANOVA?) 2. Summarize the results. (Read the entire Results section. What happened?) 3. What is a conclusion you can take from the study? (What is the take-home message from this study? How can the findings be used by non-researchers?) 4. Based on reading this study, what is a needed, future research topic? (Hint: many authors towards the end of the Discussion section talk about their view of needed future research. You can use one of theirs or a research question that is your own.)
Purchase answer to see full attachment