Question
In a compound pendulum experiment, the pendulum is made tooscillate in two different modes of oscillation. The periods ofoscillation taken are T1 and T2 respectively. Theory suggests thatT1 and T2 are related to the length of the pendulumh by the equation T2^2 - T1^2 = (4 pie^2 M L) /(3mgh) . Where m, g and L all have a fixed value provided, while Mis a constant. The whole IDEA of this question is to find out plotwhat against what to get a Straight Line graph. So, I thought thatit's plotting T2^2 - T1^2 against (1/h), but the resultant graph isnot straight. Please help? I'll give a livesaver rating... In a compound pendulum experiment, the pendulum is made tooscillate in two different modes of oscillation. The periods ofoscillation taken are T1 and T2 respectively. Theory suggests thatT1 and T2 are related to the length of the pendulumh by the equation T2^2 - T1^2 = (4 pie^2 M L) /(3mgh) . Where m, g and L all have a fixed value provided, while Mis a constant. The whole IDEA of this question is to find out plotwhat against what to get a Straight Line graph. So, I thought thatit's plotting T2^2 - T1^2 against (1/h), but the resultant graph isnot straight. Please help? I'll give a livesaver rating...
Explanation / Answer
I think your idea of what to plot was correct. Could itbe that, because of experimental error and scatter in the datapoints, you just are not seeing a perfectly straight line? Noreal-life data will give a perfectly straight line. The bestyou can do is to fit a linear function to the somewhat scattereddata. You can do this on a TI-83plus by entering x and y datain STAT tables L1 and L2, plotting data to see the scatter diagram,and then doing STAT, CALC, and either LinReg (linear regression fora straight line fit) or QuadReg (for a quadratic fit to thedata). In both cases, the calculator will give you the valueof R2 , which is a measure of the goodness of fit. Whichever fit has the largest value of R2 is the bestfit to your actual data. Hope this helps.