MIT Press

The estimates in table 2 show whether, within individuals, homeroom cortisol levels differed from baseline to the testing weeks. All columns include individual fixed effects and a quadratic of time relative to waking for a given day. The coefficient on low-stakes (high-stakes) testing approximates whether the level of cortisol differs from the baseline week to the low-stakes (high-stakes) testing week. Column 1 does not include wake time or controls for whether the sample was taken during the CAR period (15-60 minutes post-wake), as these were necessarily imputed on day 1 of each week. However, later wake times were associated with higher waking cortisol, so we added controls for wake time and CAR in column 2.^²² The homeroom estimates were similar whether controlling for these variables or not, and going forward we prefer the more conservative estimate that controls for wake time and CAR. On average, student levels of cortisol were 18 percent higher in homeroom in the high-stakes week relative to the same students’ homeroom cortisol at baseline. There was no statistical difference in cortisol in the low-stakes week relative to the baseline week, though as expected the coefficients are positive. The high-stakes and low-stakes weeks do not statistically differ from one another; future analyses should explore this relationship further.^²³

Table 2.

Changes in Level of Before-Testing Homeroom Period Cortisol by Week

	All	All	By Gender	By Poverty	By Local 911 Calls	By Ability
	(1)	(2)	(3)	(4)	(5)	(6)
Low-stakes testing	0.123	0.101	0.310^**	0.069	0.267^*	0.269^**
	(0.087)	(0.087)	(0.113)	(0.156)	(0.118)	(0.091)
High-stakes testing	0.204^**	0.176^*	0.327^**	0.295^*	0.320^**	0.273^**
	(0.075)	(0.076)	(0.119)	(0.123)	(0.121)	(0.092)
Low-stakes × female			−0.353^*
			(0.166)
High-stakes × female			−0.259⁺
			(0.150)
Low-stakes × lower poverty				0.018
				(0.188)
High-stakes × lower poverty				−0.240
				(0.164)
Low-stakes × lower crime					−0.369⁺
					(0.196)
High-stakes × lower crime					−0.266
					(0.171)
Low-stakes × higher ability						−0.329^*
						(0.163)
High-stakes × higher ability						−0.190
						(0.142)
Time of day	−0.149	0.034	0.038	0.133	0.261	0.047
	(0.612)	(0.644)	(0.646)	(0.713)	(0.721)	(0.642)
Time of day-squared	0.150	0.349	0.345	0.409	0.445	0.345
	(0.619)	(0.682)	(0.689)	(0.749)	(0.750)	(0.678)
Wake time		−0.183	−0.192	−0.184	−0.191	−0.178
		(0.139)	(0.136)	(0.138)	(0.136)	(0.141)
CAR timeframe		−0.101	−0.085	−0.151	−0.118	−0.077
		(0.175)	(0.175)	(0.194)	(0.191)	(0.175)
p(sum low-stakes testing = 0)			0.721	0.404	0.487	0.668
p(sum high-stakes testing sum = 0)			0.469	0.610	0.644	0.467
Observations	489	489	489	454	448	489
Participants	93	93	93	86	85	93

	All	All	By Gender	By Poverty	By Local 911 Calls	By Ability
	(1)	(2)	(3)	(4)	(5)	(6)
Low-stakes testing	0.123	0.101	0.310^**	0.069	0.267^*	0.269^**
	(0.087)	(0.087)	(0.113)	(0.156)	(0.118)	(0.091)
High-stakes testing	0.204^**	0.176^*	0.327^**	0.295^*	0.320^**	0.273^**
	(0.075)	(0.076)	(0.119)	(0.123)	(0.121)	(0.092)
Low-stakes × female			−0.353^*
			(0.166)
High-stakes × female			−0.259⁺
			(0.150)
Low-stakes × lower poverty				0.018
				(0.188)
High-stakes × lower poverty				−0.240
				(0.164)
Low-stakes × lower crime					−0.369⁺
					(0.196)
High-stakes × lower crime					−0.266
					(0.171)
Low-stakes × higher ability						−0.329^*
						(0.163)
High-stakes × higher ability						−0.190
						(0.142)
Time of day	−0.149	0.034	0.038	0.133	0.261	0.047
	(0.612)	(0.644)	(0.646)	(0.713)	(0.721)	(0.642)
Time of day-squared	0.150	0.349	0.345	0.409	0.445	0.345
	(0.619)	(0.682)	(0.689)	(0.749)	(0.750)	(0.678)
Wake time		−0.183	−0.192	−0.184	−0.191	−0.178
		(0.139)	(0.136)	(0.138)	(0.136)	(0.141)
CAR timeframe		−0.101	−0.085	−0.151	−0.118	−0.077
		(0.175)	(0.175)	(0.194)	(0.191)	(0.175)
p(sum low-stakes testing = 0)			0.721	0.404	0.487	0.668
p(sum high-stakes testing sum = 0)			0.469	0.610	0.644	0.467
Observations	489	489	489	454	448	489
Participants	93	93	93	86	85	93

Notes: Robust standard errors clustered by student identification. Analysis conducted at the student-day level. Outcome is the natural log of cortisol. Data come from saliva collected in homeroom. Each column represents a different regression estimate. Model limits the comparison to within-individuals, accounting for any constant observed and unobserved characteristics. Wake time is the approximate wakeup time for the day, measured with error. Column 2 is the preferred overall model. Columns 3—6 conduct the analysis by interacting the test with indicator variables for the given group. Column 4 is separated by median neighborhood poverty level (40 percent), column 5 is separated by median number of 911 calls within 0.25 of home address within a year (median = 240 calls), and column 6 is separated by median first quarter grades expressed in Z-scores (median = −0.15 standard deviations). Table includes p-values of the estimated difference in these groups for the change in cortisol for the low- and high-stakes weeks.

⁺p < 0.10; ^*p < 0.05; ^**p < 0.01.

View Large

This Feature Is Available To Subscribers Only