School and local authority characteristics associated with take-up of free school meals in Scottish secondary schools, 2014

Figures & data

Table 1. Explanation of condition and suitability ratings.

Rating	Condition	Suitability
A: Good	Performing well and operating efficiently.	Performing well and operating efficiently.
B: Satisfactory	Performing adequately but showing minor deterioration.	Performing adequately but with minor problems.
C: Poor	Showing major defects and/or not operating adequately.	Showing major problems and/or not operating optimally.
D: Bad	Economic life expired and/or risk of failure.	Does not support the delivery of services to children and communities.

Table 2. Summary statistics, parameter estimates and log-odds for the take-up of FSM adjusting for school and local authority characteristics, 342 secondary schools, Scotland, school year 2013–2014.

Variable	Mean (standard error)	Model 1 Parameter estimates (95% confidence interval) Log(take-up)	Model 2 Log-odds (95% confidence interval) take-up FSM
School characteristics
Ln(Registered & Present) (n = 345)	104.83 (3.67)	0.900 (0.822 to 0.978)	−0.556 (−.0866 to −0.246)
Ln(Registered/Roll) (n = 348)	15.96% (0.53)	0.102 (0.020 to 0.184)	0.583 (0.211 to 0.955)
Ln(Take-up Not-Registered) (n = 344)	40.90% (1.23)	0.071 (0.020 to 0.122)	0.226 (0.030 to 0.422)
Capacity (n = 362)	76.18% (0.96)	0.002 (0.000 to 0.004)	0.009 (−0.001 to 0.019)
Condition (n = 362)
A	163 (45%)	−0.086 (−0.206 to 0.034)	−0.007 (−0.440 to 0.426)
B	138 (38%)	−0.055 (−0.120 to 0.010)	−0.091 (−0.434 to 0.252)
C + D	61 (17%)
Suitability (n = 362)
A	143 (39%)	0.065 (−0.043 to 0.173)	0.138 (−0.270 to 0.546)
B	144 (40%)	0.019 (−0.030 to 0.068)	0.102 (−0.200 to 0.404)
C + D	75 (21%)
Local authority characteristics
Meal cost (n = 362)	£2.08 (0.02)	0.199 (0.015 to 0.383)	0.996 (0.190 to 1.802)
Urban per cent (n = 362)	62.20 (1.60)	0.00 (−0.006 to 0.006)	−0.001 (−0.030 to 0.028)
Rural per cent (n = 362)	22.97 (1.13)	0.002 (−0.008 to 0.012)	0.003 (−0.038 to 0.044)
Random effects
Local authority level
Variance (SE)		0.021 (0.007)	0.354 (0.118)
ICC		0.292	0.080
School level
Variance (SE)		0.051 (0.001)	0.775 (0.071)
ICC		0.708	0.175
Pupil level
^aVariance (SE)			π^2/3 = 3.289
ICC			0.746

^aFor multilevel logistic regression, the variance partition coefficient was calculated by assuming a threshold model and approximating the level 1 (pupil) variance by π²/3 = 3.29 as per Snijders and Bosker (1999).

Snijders, T. A. B., & Bosker, R. J. (1999). Multilevel analysis. London: Sage.

 Google Scholar