First 593 words of the document:
Binomial Cross-Tabulation and Chi- One Sample T-test Pearson's r and Linear Paired Samples T-test Independent Samples T-
CV Square 2 CV's MV Regression 2 MV's CV and MV test. CV and MV
Comparing one CV proportion Relationship between 2 CV's. Comparing one MV mean Relationship between 2 Relationship between CV Relationship between CV
with another CV proportion We are more likely than you with a known MV mean. Has MV's. More of this means and MV. Our this is bigger and MV. Ours is bigger
are. this changed? more of that. than our that. than yours.
Repeated Measures Independent Groups
Advertising campaign says that Men and women and Concerned that number of Are sales effected by One brand of chocolate Do stickers help sales of
20% of all chocolate bars chocolate preference. nuts was insufficient in temperature? lasts longer than another. chocolate bars?
should contain winning tickets. One sample with 2 chocolate bars. One sample with 2 scores, 1 sample two scores. 2 samples. CV and MV
One sample, CV measurements, both CV (sex One Sample, MV MV Independent groups.
n = sample size 2(df) = Value s = Std. Dev. x = before mean, s = before
Std. Dev. t (df) = t, p = Sig.
p = exact sig. p = Asymp. Sig t (df) = t, p = Sig. Predicted DV value = and
(vertical intercept + slope) x x = after mean, s = after
Nuisance Variables and Confounding factors. Nuisance IV Std. Dev
variable is any variable (not the IV) that can cause changes
Pearson's r relationship Normal looking bell
in the DV. Situational controlled by experiment design and r2 = r x r xd = test table mean,
strength. curve and within 2
Subject based controlled by randomisation. pd = test table Std. Dev
=>75 strong std. dev of mean.
Confounding happens when variable can affect both DV (report sign with values)
.45 to .74 moderate
and IV. t (df) = t, p = Sig.
.25 to .44 weak
=<.24 extremely weak
z = (value mean) / Std. Dev. Error in Methodology units of analysis
z score = proportion 2
Pearson's r tells us how Causal Relationship this causes that
IV goes in columns, DV goes much of the variation in No outliers, sub-groups and Poorly Designed selection of samples
Percentage = Proportion x 100 in rows. the DV can be explained also must be linear Logic of experiment design of study
by the linear relationship
between the IV and the Normality Normality
n = sample number, 2 = (df) = t, p = sig. DV r = -P's cor., n = #, p = sig. ( = mean, s = std.dev) ( = mean1, s = std.dev1
p = sig. (report percentages of Mean (s = std.dev.) p = lower and upper before and after litres, n = N1)
interest) t(df) = t, p = sig. (report negative values) ( d = mean diff, sd = diff) more and less
mean diff. t(df) = t, p = sig. t(df) = t, p = sig
Hypothesis/Research Question Hypothesis/Research Question Hypoth./Research Question Hypoth./Research Question Hypoth./Research Question Hypoth./Res. Question
Sample desc and % result Sample desc and % result Sample desc & result (s = sd) Relationship desc Sample desc and stats Sample desc
Significance n = , p = Relationship significance X2 Mean diff. and significance t= Pearson's r (r =) significance (before and after ) (greater and less )
95% CI if significant 95% CI if significant 95% CI if significant Conclusion = Interpretation of Mean difference d Significance (t=)
Conclusion Conclusion Conclusion Slope. Significance (t=) Conclusion
Other pages in this set
Here's a taster:
P value - p = 0.008 tells us that there is 8 in 1000 chance of getting a proportion as high/low in the sample if the proportion was still the known proportion. Conclusion = prop. has changed.
Observational = Only observes the IV, all other variables are free to vary, observes changes in DV, very prone to nuisance variables and confounding factors.
Experimental = Manipulates IV, observes changes in DV. Less prone to nuisance variables and confounding factors.…read more