Term
|
Definition
hypotheses based on multiple observations
establish reliable associations among sets of facts |
|
|
Term
|
Definition
| patterns that must occur if hypothesis is correct |
|
|
Term
| eliminates un-influential variables |
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
| logic and rational thought |
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
1. Question
2. Hypotheses
3. Derive predictions/ test consequences |
|
|
Term
|
Definition
1. Identifies what data are important to collect
2. Allows for progress in science- rejection of hypotheses |
|
|
Term
| difference between science and research |
|
Definition
| science has formal experimentation and rejection of formal hypotheses |
|
|
Term
| Is the HD method inductive or deductive? |
|
Definition
|
|
Term
| costs and benefits of induction |
|
Definition
| useful for forming hypothesis but not very applicable to scientific method |
|
|
Term
|
Definition
| evidence in support of your hypothesis |
|
|
Term
| Three types of scientific investigations (Andersen 2008) |
|
Definition
experiments
descriptive
confirmatory |
|
|
Term
| when are descriptive studies the best? |
|
Definition
| When little is known about the subject |
|
|
Term
| when are experimental studies the best? |
|
Definition
when you're building on existing knowledge,
when you need to investigate relationships, and estimate effect size |
|
|
Term
| three tenants of experimental study |
|
Definition
| control, treatment, replication |
|
|
Term
|
Definition
|
|
Term
| characteristics of a normal distribution |
|
Definition
bell shaped, 50% lie on either side of the mean
mean and median are same |
|
|
Term
|
Definition
low SD means values are close to the mean, high SD means the values are spread out
The smaller the standard deviation the more concentrated the data |
|
|
Term
|
Definition
significance level
the probability used as criterion for rejection of the null |
|
|
Term
| normally distributed data, alpha is |
|
Definition
| 0.05, which is 1.96 (two stdevs from mean) |
|
|
Term
|
Definition
| testing if value falls either to left or right of mean |
|
|
Term
|
Definition
| testing if value falls within either tail |
|
|
Term
| How do you deal with non-normally distributed variables? |
|
Definition
parametric tests
non-parametric tests
transformations |
|
|
Term
|
Definition
| rejection of a null when it is true |
|
|
Term
|
Definition
accept the null and draw erroneous conclusion
beta error |
|
|
Term
| power of a stat test is defined as |
|
Definition
| 1-beta: the prob of correctly rejecting the null when it is false |
|
|
Term
|
Definition
larger difference between the mean of the data and null mean
larger for larger alpha
larger for larger pop
larger for smaller variances of sample mean |
|
|
Term
|
Definition
1.State null and alternative-BEFORE DATA COLLECTION
2.Declare level of significance
3.Collect data
4.Perform statistical tests
5.Get critical value, d.f., and p value for tests
6.State confidence limits
7.Assess null and alternative based on critical values, d.f. and p value
8.State conclusions in terms of biological or other practical significance. |
|
|
Term
|
Definition
establish research hypotheses about the fact giving processes driving nature
basis for new hypotheses
post hoc
looking at all data |
|
|
Term
|
Definition
| process of gaining reliable knowledge through experimentation and rejection of tested hypotheses |
|
|
Term
|
Definition
| parameter index matrix (how to handle parameters) |
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
| Brownie et all adjust PIM for juviniles |
|
Definition
1234 5234
234 634
34 74
4 8 |
|
|
Term
|
Definition
>3
most likely violating an assumption |
|
|
Term
|
Definition
<1
most likely violating an assumption |
|
|
Term
|
Definition
how well the data meets the assumptions
have to calculate c hat
use most paramaterized model |
|
|
Term
| only use Goodness of fit for ___? |
|
Definition
|
|
Term
|
Definition
| test> bootstrap GOF> chat etc. > 100 simulations |
|
|
Term
|
Definition
| quasi- adjusts AIC according to c hat if it's about 3. for a problem in your data ...non independent parameters... overdispersion...(non normal distribution) |
|
|
Term
|
Definition
|
|
Term
| multi model inference (MMI)definition |
|
Definition
| making inductive inferences from ALL a priori models |
|
|
Term
|
Definition
model averaging
unconditional variance
gauging relative importance of variance
confidence sets on models |
|
|
Term
|
Definition
process of obtaining a weighted average of the estimates using all models
can be done for prediction or for parameters |
|
|
Term
| predictive model averaging |
|
Definition
multiply each prediction (y hat) by its weight and sum them= survival across all models
Parameter= DEPENDENT varible [Y]= B0+B1 |
|
|
Term
| parameter model averaging |
|
Definition
model estimate for the parameter (SAS output) and multiply by AICw. Sum for each parameter. Those values are what you use in the equation.
Parameter=INDEPENDENT variable Y=[B0+B1] |
|
|
Term
|
Definition
| sampling variation due to model selection uncertainty (per model) |
|
|
Term
|
Definition
| variance for all models, sum of model variance and weighted variance |
|
|
Term
| What do you report for AIC? |
|
Definition
1. top table (AIC, AICc, wi,...)
2. variables of importance
3. model averaged parameters
4. demonstrate variance/ parameter |
|
|
Term
| Capture Mark Recapture Theory |
|
Definition
any sample of animals that is encountered is an incomplete count of the population
N=n/phat or, true pop=sample/capture prob. |
|
|
Term
| Generalizations of Capture Mark Recapture |
|
Definition
marks aren't just tags
Mark doesn't affect animal
capture probability,phat assumed = over all space and time |
|
|
Term
|
Definition
| hunting/ fishing situations (case 1) |
|
|
