Professional Documents
Culture Documents
Report - Statis Ti Cs /analysis: Template
Report - Statis Ti Cs /analysis: Template
TEMPLATE
STATISTICS
Talk about amount of data collected, quote quantities. Systematic errors if any and what was done to remedy this. Table of stats including: Number of data sets total Sample size arithmetic mean standard error of the mean uncertainty Confidence interval 90% 99%
Talk about standard deviations of specific data sets, skew of data, possible anomalies and their effect on the overall data set. Quantitative or qualitative fit of data to accepted or theoretical values mentioned if applicable. For last experiment correlation and covariance of variables.
ANALYSIS
Mistakes we made during the experiment, and therefore what we could have done next time. Better methods of collecting data with hindsight. What actually went well with the experiment. What skills we learned e.g. excel functions, statistical analysis, interpersonal skills, time management etc.
1.41%
TA B L E 1 : B A S I C D E S C R I P TI V E S TA TI S TI C S A B O U T T HE D I F F R A C T I O N G R A T I N G S E X P E R I M E N T.
The actual labelled value given by the laser for the wavelength was 532 nm, 0.27% lower than the calculated value, which seems to be quite accurate. The skew of the data was 1.60 which means that more values were larger than the mean than were smaller than it. There was quite a small standard deviation of 7.52 nm. There was a mysterious cluster, four consequetive points, of data that were between three and four standard deviations away from the mean, all skewed in the same direction, which implies some kind of unaccounted for systematic error. It should however be noted that later data sets were considerably less accurate that sets recorded first because of the higher fractional error due to a lower number of visible maximas observed.
12.52%
TA B L E 2 : B A S I C D E S C R I P TI V E S TA TI S TI C S A B O U T T HE S I N G L E S L I T E X P E R I M E N T .
The standard deviation was 7.85 m, rather high as is shown in the uncertainty. This was down the large fractional errors for small input data, which combined in quadrature. The slit to screen distance should have been larger to combat this. The skew was 2.95, caused by the high error first maxima measurements, one of which was more than three standard deviations away from the weighted mean, quite odd for such a small sample size.
As can be seen in Chart 3, predictions of maxima positions using the weighted mean closely agree with observed the positions observed. It should again be stressed that there is no simple model that predicts maxima positions for single slit diffraction.
uncertainty
9.00%
TA B L E 3 : B A S I C D E S C R I P TI V E S TA TI S TI C S A B O U T T HE D O U B L E S L I T E X P E R I M E N T .
There was a standard deviation of 18.0 m, and a skew of -0.313. There were two results which were around four standard deviations away from the mean, one high and one low; this is plausibly down to random error with so large a sample size. There was little agreement between different data sets, even the last two which coincidently were measured under (supposedly) the exact same conditions. This implies there was some variable which was not controlled.
It is noteworthy that other physics groups reported calculating similar values for the slit separation. There was some agreement between experimental data and theoretical predictions assuming that the separation was the average, mostly in sets two and three. Sets one and four did not look like they fit the prediction much at all. See chart five (in appendix, too big to fit into this section).
uncertainty 21.4 %
TA B L E 4 : B A S I C D E S C R I P TI V E S TA TI S TI C S A B O U T T HE A P E R TU R E D I F F R A C TI O N EXPERIMENT.
It was also assumed that the circular aperture would theoretically produce a pattern similar to the single slit, but in a continuous circle. (need to test)!!!!!!!!
uncertainty 29.1 %
TA B L E 5 : B A S I C D E S C R I P TI V E S TA TI S TI C S A B O U T T HE W I R E D I F F R A C T I O N E X P E R I M E N T .
The results had a large standard deviation of 27.4 m The wire width calculations were considerably further from the mean when calculated from the first maxima readings of each data set, this is because fractional errors add in quadrature and the fractional errors are much greater in small measurements. This greatly increased the standard deviation.